SlideShare a Scribd company logo
1 of 32
Download to read offline
Motivation
                       Method
                        Results
                      Summary




                        GPU-Euler
            Sequence Assembly using GPGPU


              S. Mahmood          H. Rangwala

                Department of Computer Science
                   George Mason University


International Conference on High Performance Computing &
                   Communications, 2011
                       Ban, Canada



             Mahmood, Rangwala    GPU-Euler
Motivation
                              Method
                               Results
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                         Genome Assembly
                              Method
                                         Previous Work
                               Results
                                         GPGPU
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                           Genome Assembly
                                Method
                                           Previous Work
                                 Results
                                           GPGPU
                               Summary



Genome


     Genome  a biological blueprint.
     Very long chains of four types of
     nucleobases.
         Adenine
         Guanine
         Cytosine
         Thymine
     Important to understand the
                                               Figure: Double Helix DNA
     function of the organism.
                                               representation
 1
    1 Image courtesy of Image Library of Biological Macromolecules, Jena,
 Germany. http://www.imb-jena.de/IMAGE.html
                      Mahmood, Rangwala    GPU-Euler
Motivation
                                            Genome Assembly
                                 Method
                                            Previous Work
                                  Results
                                            GPGPU
                                Summary



Sequence Assembly
Challenges

         Total number of nucleobases in a genome is very large
             eg. Human Genome has 3.2 Billion base pairs.
         Existing technologies can only read a fraction of this long
         strand.
         Smaller fragments(reads) are required to be stitched together.




                          Figure: Sequence Assembly

   2
     2 Image courtesy of Center for BioInformatics Computational Biology,UMD.
   www.cbcb.umd.edu/research/assembly_primer.html
                       Mahmood, Rangwala    GPU-Euler
Motivation
                                        Genome Assembly
                             Method
                                        Previous Work
                              Results
                                        GPGPU
                            Summary



Problem Statement



     Given a set of alphabets ∑ = {A, G , C , T } and a set of strings
     R = {r1 , r2 , r3 . . . rn } over alphabet Σ

     Construct Super String S, containing all the strings from R.
     Similar to Shortest Common Super string.
     Need to consider Repeats.
     Massive volume of data




                   Mahmood, Rangwala    GPU-Euler
Motivation
                                         Genome Assembly
                              Method
                                         Previous Work
                               Results
                                         GPGPU
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                           Genome Assembly
                                Method
                                           Previous Work
                                 Results
                                           GPGPU
                               Summary



Sequence Assembly
Techniques  Tools




        Greedy Assemblers
             VCAKE
        Overlap-layout-consensus
             Celera
        Eulerian Path
             Euler, EulerSR, Velvet
             ABySS




                      Mahmood, Rangwala    GPU-Euler
Motivation
                                         Genome Assembly
                              Method
                                         Previous Work
                               Results
                                         GPGPU
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                              Genome Assembly
                                   Method
                                              Previous Work
                                    Results
                                              GPGPU
                                  Summary



General Purpose GPU Computing

      GPUs for General Purpose Computing
      Massive parallelism for application.
      nVidia CUDA , a framework for
      development on nVidia GPUs.
      Similar model for parallel computation
           Parallel Random Access Machine
           (PRAM)                                         Figure: CUDA Application
           Single Instruction Multiple Data               Stack
           (SIMD)
  3



      3 Image   courtesy of nVidia : nVidia CUDA Toolkit Reference Manual
                         Mahmood, Rangwala    GPU-Euler
Motivation
                                               Genome Assembly
                                    Method
                                               Previous Work
                                     Results
                                               GPGPU
                                   Summary



CUDA
Compute Unied Device Architecture



A CUDA enabled device has
       Symmetric Multiprocessor (SM)
       Each SM has a set of Streaming
       Processors (SP).
       Global Memory.
       Concurrent execution of same
       code on all SM.
       Computations use GPU memory.
                                                     Figure: Hardware Architecture
   4
       4 Image   courtesy of nVidia : nVidia CUDA Toolkit Reference Manual
                          Mahmood, Rangwala    GPU-Euler
Motivation
                                         Parallel Eulerian Assembly
                              Method
                                         Time Complexity Analysis
                               Results
                                         Evaluation
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                        Parallel Eulerian Assembly
                             Method
                                        Time Complexity Analysis
                              Results
                                        Evaluation
                            Summary



Concepts

de-Bruijn Graph A directed graph
                 vertice are k length
                 word
                 edge represents a
                 k − 1 between
                 vertices.
    Contigs Assembled sequences from
            the input data.
  EulerTour A graph traversal visiting                Figure: de-Bruijn Graph
            each edge only once.


                   Mahmood, Rangwala    GPU-Euler
Motivation
                                        Parallel Eulerian Assembly
                             Method
                                        Time Complexity Analysis
                              Results
                                        Evaluation
                            Summary



Objective




      Represent input read as a de-Bruijn graph
      Each edge would correspond to a single base.
      An Euler tour will visit each base only once.




                   Mahmood, Rangwala    GPU-Euler
Motivation
                                                Parallel Eulerian Assembly
                              Method
                                                Time Complexity Analysis
                               Results
                                                Evaluation
                             Summary



Parallel Eulerian Assembly


                                           FASTA file
                                             (input)




                                             Reads




  Construct de-Bruijn Graph              Debruijn Graph
                                          Construction    Graph
                                                                   Euler Tour
                                                                  Construction   Annotated Graph
                                                                                                   Identify Contigs




  Find Euler Tour                                                 EulerGPU
                                                                                                      Contigs



  Output Contigs                                                                                     FASTA file
                                                                                                      (output)




                                                     Figure: GPU Euler Work ow




                    Mahmood, Rangwala           GPU-Euler
Motivation
                                        Parallel Eulerian Assembly
                             Method
                                        Time Complexity Analysis
                              Results
                                        Evaluation
                            Summary



Parallel de-Bruijn Graph Construction

                                                                        FASTA file


                                          CUDA

   Assign each CUDA thread to              Count
                                           Edges

   one read.
                                                             CUDA
   Generate k -mers and                                       Setup
                                                             Vertices
   k + 1-mers.

   Store them in a hash table.                                                       CUDA
                                                                                      Setup

   Create nodes from k-mers                                                           Edges


   and vertices from k + 1-mers.

                                              Figure: Graph Construction


                   Mahmood, Rangwala    GPU-Euler
Motivation
                                          Parallel Eulerian Assembly
                               Method
                                          Time Complexity Analysis
                                Results
                                          Evaluation
                              Summary



Parallel Euler Tour

    Create a Edge Successor Graph
    from de-Bruijn Graph.                                CUDA


    Identify Circuits in the Edge
                                                          Assign             DeBruijn Graph
                                                        Successor




    Successor Graph                                    Annotated Graph




    Create a Circuit Graph by
                                                                                              CUDA
                                                                                               Create
                                                                                               Circuit
                                                    CUDA


    identifying adjacent circuits.
                                                                         Comp. Label           Graph
                                                      Find
                                                   Component
                                                                                               Circuit Graph



    Calculate a spanning tree for                                                                  Find


    Circuit.
                                                                          Spanning Tree        Spanning Tree
                                                   CUDA
                                                   Execute
                                                    Swipe
                                                                                               Euler Tour

    Traverse Circuit Graph and
    switch successor edges of                       Figure: Parallel Euler Tour
    adjacent Circuits.


                     Mahmood, Rangwala    GPU-Euler
Motivation
                                      Parallel Eulerian Assembly
                           Method
                                      Time Complexity Analysis
                            Results
                                      Evaluation
                          Summary



Decomposition of Dierent Phases

        Phase                                       Computation
        I/O and k-mer Extraction                    CPU + GPU
        Hash Table Construction                       GPU
        debruijn Graph Construction                   GPU
        Euler Tour Construction                     GPU + CPU
        Sub-steps for Euler Tour Construction
          Finding Connected Component                      GPU
          Circuit Graph Creation                           GPU
          Spanning Tree                                    CPU
          Swipe Execution                                  GPU
          Traversal (Other)                                GPU
        Contig Generation (O/P)                            CPU


                 Mahmood, Rangwala    GPU-Euler
Motivation
                                         Parallel Eulerian Assembly
                              Method
                                         Time Complexity Analysis
                               Results
                                         Evaluation
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                      Parallel Eulerian Assembly
                           Method
                                      Time Complexity Analysis
                            Results
                                      Evaluation
                          Summary



Time Complexity Analysis



                  Step                   Complexity           Processors
      de-Bruijn Graph Construction          O (1)               O (n )
         Euler Tour Construction           (log n)              O (n )
              Spanning Tree              O (log |V |)          O (|V |)
                GPU-Euler                 O (log n )            O (n )




                 Mahmood, Rangwala    GPU-Euler
Motivation
                                         Parallel Eulerian Assembly
                              Method
                                         Time Complexity Analysis
                               Results
                                         Evaluation
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                       Parallel Eulerian Assembly
                            Method
                                       Time Complexity Analysis
                             Results
                                       Evaluation
                           Summary



Experimental Protocol


      Compared Timing, N50 Score, Mean length with EulerSR
      using various parameters.
          Why EulerSR
              Based on same concept
              Shared memory approach
              Support short reads
      Contigs with length  100 were included in the comparison.
      Calculated contig converge using MUMMER.
      Individual GPU Computations were timed as well.



                  Mahmood, Rangwala    GPU-Euler
Motivation
                              Method     Data sets
                               Results   Results
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                  Method     Data sets
                                   Results   Results
                                 Summary



Data Sets
Genome size and number of simulated reads for dierent read length




     Genome                       Length          36 bp      50 bp    256 bp
     Campylobacter Jejuni        1,641,481       911,934    656,593   128,241
     Neisseria Meningitidis      2,184,406      1,213,559   873,763   170,657
     Lactococcus Lactisd         2,635,589      1,314,216   946,236   184,812




                        Mahmood, Rangwala    GPU-Euler
Motivation
                              Method     Data sets
                               Results   Results
                             Summary



Outline

  1   Motivation
       Genome Assembly
       Previous Work
       GPGPU

  2   Method
       Parallel Eulerian Assembly
       Time Complexity Analysis
       Evaluation

  3   Results
        Data sets
        Results


                    Mahmood, Rangwala    GPU-Euler
Motivation
                                                                                                            Method                                    Data sets
                                                                                                              Results                                 Results
                                                                                                          Summary



Results
Execution Time Comparison

                                               Runtime Comparison for Campylobacter Jejuni                                                                                                                    Runtime Comparison for Neisseria Meningitidis
                                      180                                                                                                                                           300
                                                                                                                     EulerSR                                                                                                                                       EulerSR
                                                                                                                    EulerSR*                                                                                                                                      EulerSR*
                                                                                                                   GPU-Euler                                                                                                                                     GPU-Euler
                                      160
                                                                                                                                                                                    250

                                      140
            Assembly Time (seconds)




                                                                                                                                                          Assembly Time (seconds)
                                                                                                                                                                                    200
                                      120


                                      100                                                                                                                                           150


                                      80
                                                                                                                                                                                    100

                                      60

                                                                                                                                                                                    50
                                      40


                                      20                                                                                                                                             0
                                            36bp                  50bp                                     256bp                                                                                         36bp                     50bp                   256bp
                                                              Read Length                                                                                                                                                     Read Length




                                                                                                                                 Runtime Comparison for Lactococcus Lactis
                                                                                                   300
                                                                                                                                                                                                    EulerSR
                                                                                                                                                                                                   EulerSR*
                                                                                                                                                                                                  GPU-Euler

                                                                                                   250
                                                                         Assembly Time (seconds)




                                                                                                   200



                                                                                                   150



                                                                                                   100



                                                                                                   50



                                                                                                    0
                                                                                                                          36bp                     50bp                                   256bp
                                                                                                                                               Read Length




                                                     Mahmood, Rangwala                                                                                GPU-Euler
Motivation
                                                                                               Method                              Data sets
                                                                                                Results                            Results
                                                                                              Summary



Results
N50 Score Comparison

                                          N50 Score Comparison for Campylobacter Jejuni                                                                                             N50 Score Comparison for Neisseria Meningitidis
                                120000                                                                                                                      35000
                                                                                                    EulerSR                                                                                                                             EulerSR
                                                                                                   EulerSR*                                                                                                                            EulerSR*
                                                                                                  GPU-Euler                                                                                                                           GPU-Euler
                                                                                                                                                            30000
                                100000


                                                                                                                                                            25000
                                80000
            N50 Score (bases)




                                                                                                                                        N50 Score (bases)
                                                                                                                                                            20000

                                60000

                                                                                                                                                            15000

                                40000
                                                                                                                                                            10000


                                20000
                                                                                                                                                             5000



                                     0                                                                                                                          0
                                         36bp                 50bp                            256bp                                                                             36bp                     50bp                  256bp
                                                          Read Length                                                                                                                                Read Length




                                                                                                              N50 Score Comparison for Lactococcus Lactis
                                                                                      80000
                                                                                                                                                                          EulerSR
                                                                                                                                                                         EulerSR*
                                                                                                                                                                        GPU-Euler
                                                                                      70000


                                                                                      60000
                                                                  N50 Score (bases)




                                                                                      50000


                                                                                      40000


                                                                                      30000


                                                                                      20000


                                                                                      10000


                                                                                          0
                                                                                                          36bp                   50bp                               256bp
                                                                                                                             Read Length




                                                Mahmood, Rangwala                                                                  GPU-Euler
Motivation
                                                                                         Method                         Data sets
                                                                                           Results                      Results
                                                                                       Summary



Results
Accuracy Comparison

                                              Campylobacter Jejuni                                                                                                         Neisseria Meningitidis
                                100                                                                                                            100
                                                                                                  EulerSR                                                                                                     EulerSR
                                                                                                 EulerSR*                                                                                                    EulerSR*
                                                                                                GPU-Euler                                      90                                                           GPU-Euler


                                95                                                                                                             80

                                                                                                                                               70
            Weighted Accuracy




                                                                                                                           Weighted Accuracy
                                90                                                                                                             60

                                                                                                                                               50

                                85                                                                                                             40

                                                                                                                                               30

                                80                                                                                                             20

                                                                                                                                               10

                                75                                                                                                              0
                                      36bp           50bp                               256bp                                                                       36bp           50bp             256bp
                                                  Read Length                                                                                                                  Read Length




                                                                                                              Lactococcus Lactis
                                                                                100
                                                                                                                                                               EulerSR
                                                                                                                                                              EulerSR*
                                                                                                                                                             GPU-Euler


                                                                                95
                                                            Weighted Accuracy




                                                                                90




                                                                                85




                                                                                80




                                                                                75
                                                                                                       36bp         50bp                             256bp
                                                                                                                Read Length




                                             Mahmood, Rangwala                                                          GPU-Euler
Motivation
                             Method     Data sets
                              Results   Results
                            Summary



GPU Euler Phase Distribution


   Phase                                          Computation     % Time
   I/O and k-mer Extraction                       CPU + GPU     77.29+1.44
   Hash Table Construction                          GPU             0.31
   debruijn Graph Construction                      GPU             1.15
   Euler Tour Construction                        GPU + CPU
   Sub-steps for Euler Tour Construction
     Finding Connected Component                     GPU          10.06
     Spanning Tree                                   CPU           0.06
     Swipe Execution                                 GPU           0.01
     Circuit Graph  Traversal (Other)               GPU           0.72
   Contig Generation (O/P)                           CPU           4.39


                   Mahmood, Rangwala    GPU-Euler
Motivation
                            Method
                             Results
                           Summary



Summary



    Exploiting GPUs for Sequence Assembly.
    Implementation of PRAM algorithm on CUDA devices.


    Outlook
          No Error Correction
          Graph Simplication.




                  Mahmood, Rangwala    GPU-Euler
Motivation
                      Method
                       Results
                     Summary



Questions




                        Questions?




            Mahmood, Rangwala    GPU-Euler
Motivation
                      Method
                       Results
                     Summary



Thank you




                       Thank you!!




            Mahmood, Rangwala    GPU-Euler

More Related Content

What's hot

Final report
Final reportFinal report
Final reportUC Davis
 
NEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODING
NEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODINGNEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODING
NEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODINGcscpconf
 
IRJET - Implementation of Neural Network on FPGA
IRJET - Implementation of Neural Network on FPGAIRJET - Implementation of Neural Network on FPGA
IRJET - Implementation of Neural Network on FPGAIRJET Journal
 
Paper on experimental setup for verifying - "Slow Learners are Fast"
Paper  on experimental setup for verifying  - "Slow Learners are Fast"Paper  on experimental setup for verifying  - "Slow Learners are Fast"
Paper on experimental setup for verifying - "Slow Learners are Fast"Robin Srivastava
 
An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...
An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...
An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...IJCSIS Research Publications
 
PCA Clouds (ICET 2005)
PCA Clouds (ICET 2005)PCA Clouds (ICET 2005)
PCA Clouds (ICET 2005)IT Industry
 
Median based parallel steering kernel regression for image reconstruction
Median based parallel steering kernel regression for image reconstructionMedian based parallel steering kernel regression for image reconstruction
Median based parallel steering kernel regression for image reconstructioncsandit
 
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTIONMEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTIONcsandit
 
Enhance similarity searching algorithm with optimized fast population count m...
Enhance similarity searching algorithm with optimized fast population count m...Enhance similarity searching algorithm with optimized fast population count m...
Enhance similarity searching algorithm with optimized fast population count m...IOSR Journals
 
ANN based STLF of Power System
ANN based STLF of Power SystemANN based STLF of Power System
ANN based STLF of Power SystemYousuf Khan
 
Object Centric Reflection
Object Centric ReflectionObject Centric Reflection
Object Centric ReflectionESUG
 
A COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTION
A COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTIONA COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTION
A COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTIONIJCSEA Journal
 
M sc thesis of nicolo' savioli
M sc thesis of nicolo' savioliM sc thesis of nicolo' savioli
M sc thesis of nicolo' savioliNicolò Savioli
 
0 eeg based 3 d visual fatigue evaluation using cnn
0 eeg based 3 d visual fatigue evaluation using cnn0 eeg based 3 d visual fatigue evaluation using cnn
0 eeg based 3 d visual fatigue evaluation using cnnHoopeer Hoopeer
 

What's hot (18)

Final report
Final reportFinal report
Final report
 
NEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODING
NEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODINGNEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODING
NEW IMPROVED 2D SVD BASED ALGORITHM FOR VIDEO CODING
 
238 243
238 243238 243
238 243
 
IRJET - Implementation of Neural Network on FPGA
IRJET - Implementation of Neural Network on FPGAIRJET - Implementation of Neural Network on FPGA
IRJET - Implementation of Neural Network on FPGA
 
I017425763
I017425763I017425763
I017425763
 
Nvidia GTC 2014 Talk
Nvidia GTC 2014 TalkNvidia GTC 2014 Talk
Nvidia GTC 2014 Talk
 
Paper on experimental setup for verifying - "Slow Learners are Fast"
Paper  on experimental setup for verifying  - "Slow Learners are Fast"Paper  on experimental setup for verifying  - "Slow Learners are Fast"
Paper on experimental setup for verifying - "Slow Learners are Fast"
 
An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...
An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...
An Efficient APOA Techniques For Generalized Residual Vector Quantization Bas...
 
20120140505010
2012014050501020120140505010
20120140505010
 
PCA Clouds (ICET 2005)
PCA Clouds (ICET 2005)PCA Clouds (ICET 2005)
PCA Clouds (ICET 2005)
 
Median based parallel steering kernel regression for image reconstruction
Median based parallel steering kernel regression for image reconstructionMedian based parallel steering kernel regression for image reconstruction
Median based parallel steering kernel regression for image reconstruction
 
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTIONMEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
 
Enhance similarity searching algorithm with optimized fast population count m...
Enhance similarity searching algorithm with optimized fast population count m...Enhance similarity searching algorithm with optimized fast population count m...
Enhance similarity searching algorithm with optimized fast population count m...
 
ANN based STLF of Power System
ANN based STLF of Power SystemANN based STLF of Power System
ANN based STLF of Power System
 
Object Centric Reflection
Object Centric ReflectionObject Centric Reflection
Object Centric Reflection
 
A COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTION
A COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTIONA COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTION
A COMPARATIVE STUDY OF BACKPROPAGATION ALGORITHMS IN FINANCIAL PREDICTION
 
M sc thesis of nicolo' savioli
M sc thesis of nicolo' savioliM sc thesis of nicolo' savioli
M sc thesis of nicolo' savioli
 
0 eeg based 3 d visual fatigue evaluation using cnn
0 eeg based 3 d visual fatigue evaluation using cnn0 eeg based 3 d visual fatigue evaluation using cnn
0 eeg based 3 d visual fatigue evaluation using cnn
 

Viewers also liked

Eulermethod3
Eulermethod3Eulermethod3
Eulermethod3stellajoh
 
Euler method notes sb
Euler method notes sbEuler method notes sb
Euler method notes sbjanetvmiller
 
AP Calculus AB March 25, 2009
AP Calculus AB March 25, 2009AP Calculus AB March 25, 2009
AP Calculus AB March 25, 2009Darren Kuropatwa
 
Introduction to Numerical Methods for Differential Equations
Introduction to Numerical Methods for Differential EquationsIntroduction to Numerical Methods for Differential Equations
Introduction to Numerical Methods for Differential Equationsmatthew_henderson
 
Successive iteration method for reconstruction of missing data
Successive iteration method for reconstruction of missing dataSuccessive iteration method for reconstruction of missing data
Successive iteration method for reconstruction of missing dataIAEME Publication
 
Calculus II - 13
Calculus II - 13Calculus II - 13
Calculus II - 13David Mao
 
Perturbation methods last
Perturbation methods lastPerturbation methods last
Perturbation methods lastSabir Ahmed
 
Euler Method using MATLAB
Euler Method using MATLABEuler Method using MATLAB
Euler Method using MATLABM.Ahmed Waqas
 
Automobile 3rd sem aem ppt.2016
Automobile 3rd sem aem ppt.2016Automobile 3rd sem aem ppt.2016
Automobile 3rd sem aem ppt.2016kalpeshvaghdodiya
 
metode euler
metode eulermetode euler
metode eulerRuth Dian
 
Interactives Methods
Interactives MethodsInteractives Methods
Interactives MethodsUIS
 
Numerical Methods
Numerical MethodsNumerical Methods
Numerical MethodsESUG
 
AP Calculus AB March 26, 2009
AP Calculus AB March 26, 2009AP Calculus AB March 26, 2009
AP Calculus AB March 26, 2009Darren Kuropatwa
 
Ordinary differential equation
Ordinary differential equationOrdinary differential equation
Ordinary differential equationJUGAL BORAH
 

Viewers also liked (20)

Eulermethod3
Eulermethod3Eulermethod3
Eulermethod3
 
Euler method notes sb
Euler method notes sbEuler method notes sb
Euler method notes sb
 
Day 4 examples
Day 4 examplesDay 4 examples
Day 4 examples
 
10 09
10 0910 09
10 09
 
DLR_DG_AZIZ_2003
DLR_DG_AZIZ_2003DLR_DG_AZIZ_2003
DLR_DG_AZIZ_2003
 
AP Calculus AB March 25, 2009
AP Calculus AB March 25, 2009AP Calculus AB March 25, 2009
AP Calculus AB March 25, 2009
 
Introduction to Numerical Methods for Differential Equations
Introduction to Numerical Methods for Differential EquationsIntroduction to Numerical Methods for Differential Equations
Introduction to Numerical Methods for Differential Equations
 
Successive iteration method for reconstruction of missing data
Successive iteration method for reconstruction of missing dataSuccessive iteration method for reconstruction of missing data
Successive iteration method for reconstruction of missing data
 
Calculus II - 13
Calculus II - 13Calculus II - 13
Calculus II - 13
 
Perturbation methods last
Perturbation methods lastPerturbation methods last
Perturbation methods last
 
Euler Method using MATLAB
Euler Method using MATLABEuler Method using MATLAB
Euler Method using MATLAB
 
Automobile 3rd sem aem ppt.2016
Automobile 3rd sem aem ppt.2016Automobile 3rd sem aem ppt.2016
Automobile 3rd sem aem ppt.2016
 
Differential equations intro
Differential equations introDifferential equations intro
Differential equations intro
 
metode euler
metode eulermetode euler
metode euler
 
Interactives Methods
Interactives MethodsInteractives Methods
Interactives Methods
 
Ch02 7
Ch02 7Ch02 7
Ch02 7
 
Numerical Methods
Numerical MethodsNumerical Methods
Numerical Methods
 
AP Calculus AB March 26, 2009
AP Calculus AB March 26, 2009AP Calculus AB March 26, 2009
AP Calculus AB March 26, 2009
 
Calc 6.1b
Calc 6.1bCalc 6.1b
Calc 6.1b
 
Ordinary differential equation
Ordinary differential equationOrdinary differential equation
Ordinary differential equation
 

Similar to Hpcc euler

IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...
IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...
IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...IRJET Journal
 
A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...
A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...
A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...ijdms
 
A Survey of Spiking Neural Networks and Support Vector Machine Performance By...
A Survey of Spiking Neural Networks and Support Vector Machine Performance By...A Survey of Spiking Neural Networks and Support Vector Machine Performance By...
A Survey of Spiking Neural Networks and Support Vector Machine Performance By...ijsc
 
4213ijaia02
4213ijaia024213ijaia02
4213ijaia02ijaia
 
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTIONMEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTIONcscpconf
 
Training artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithmTraining artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithmA. Roy
 
Training artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithmTraining artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithmA. Roy
 
SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY
SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY
SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY cscpconf
 
An Application of Genetic Programming for Power System Planning and Operation
An Application of Genetic Programming for Power System Planning and OperationAn Application of Genetic Programming for Power System Planning and Operation
An Application of Genetic Programming for Power System Planning and OperationIDES Editor
 
Evolutionary Testing Approach for Solving Path- Oriented Multivariate Problems
Evolutionary Testing Approach for Solving Path- Oriented Multivariate ProblemsEvolutionary Testing Approach for Solving Path- Oriented Multivariate Problems
Evolutionary Testing Approach for Solving Path- Oriented Multivariate ProblemsIDES Editor
 
Automated-tuned hyper-parameter deep neural network by using arithmetic optim...
Automated-tuned hyper-parameter deep neural network by using arithmetic optim...Automated-tuned hyper-parameter deep neural network by using arithmetic optim...
Automated-tuned hyper-parameter deep neural network by using arithmetic optim...IJECEIAES
 
Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...
Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...
Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...TELKOMNIKA JOURNAL
 
Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...
Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...
Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...ijcax
 

Similar to Hpcc euler (15)

Thesis Defense
Thesis DefenseThesis Defense
Thesis Defense
 
IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...
IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...
IRJET-Performance Enhancement in Machine Learning System using Hybrid Bee Col...
 
A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...
A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...
A SURVEY OF SPIKING NEURAL NETWORKS AND SUPPORT VECTOR MACHINE PERFORMANCE BY...
 
A Survey of Spiking Neural Networks and Support Vector Machine Performance By...
A Survey of Spiking Neural Networks and Support Vector Machine Performance By...A Survey of Spiking Neural Networks and Support Vector Machine Performance By...
A Survey of Spiking Neural Networks and Support Vector Machine Performance By...
 
cuTau Leaping
cuTau LeapingcuTau Leaping
cuTau Leaping
 
4213ijaia02
4213ijaia024213ijaia02
4213ijaia02
 
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTIONMEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
MEDIAN BASED PARALLEL STEERING KERNEL REGRESSION FOR IMAGE RECONSTRUCTION
 
Training artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithmTraining artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithm
 
Training artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithmTraining artificial neural network using particle swarm optimization algorithm
Training artificial neural network using particle swarm optimization algorithm
 
SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY
SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY
SPEED-UP IMPROVEMENT USING PARALLEL APPROACH IN IMAGE STEGANOGRAPHY
 
An Application of Genetic Programming for Power System Planning and Operation
An Application of Genetic Programming for Power System Planning and OperationAn Application of Genetic Programming for Power System Planning and Operation
An Application of Genetic Programming for Power System Planning and Operation
 
Evolutionary Testing Approach for Solving Path- Oriented Multivariate Problems
Evolutionary Testing Approach for Solving Path- Oriented Multivariate ProblemsEvolutionary Testing Approach for Solving Path- Oriented Multivariate Problems
Evolutionary Testing Approach for Solving Path- Oriented Multivariate Problems
 
Automated-tuned hyper-parameter deep neural network by using arithmetic optim...
Automated-tuned hyper-parameter deep neural network by using arithmetic optim...Automated-tuned hyper-parameter deep neural network by using arithmetic optim...
Automated-tuned hyper-parameter deep neural network by using arithmetic optim...
 
Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...
Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...
Fine-grained or coarse-grained? Strategies for implementing parallel genetic ...
 
Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...
Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...
Cell Charge Approximation for Accelerating Molecular Simulation on CUDA-Enabl...
 

Recently uploaded

Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsMark Billinghurst
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Patryk Bandurski
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxhariprasad279825
 
DevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache MavenDevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache MavenHervé Boutemy
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupFlorian Wilhelm
 
SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024Lorenzo Miniero
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clashcharlottematthew16
 
Dev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio WebDev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio WebUiPathCommunity
 
What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024Stephanie Beckett
 
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostLeverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostZilliz
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr LapshynFwdays
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Scott Keck-Warren
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsMemoori
 
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfGen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfAddepto
 
My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024The Digital Insurer
 
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding ClubUnleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding ClubKalema Edgar
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfAlex Barbosa Coqueiro
 
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLDeveloper Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLScyllaDB
 
Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 3652toLead Limited
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationSafe Software
 

Recently uploaded (20)

Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR Systems
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptx
 
DevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache MavenDevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache Maven
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project Setup
 
SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clash
 
Dev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio WebDev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio Web
 
What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024
 
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostLeverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial Buildings
 
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfGen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdf
 
My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024
 
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding ClubUnleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding Club
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdf
 
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQLDeveloper Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQL
 
Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
 

Hpcc euler

  • 1. Motivation Method Results Summary GPU-Euler Sequence Assembly using GPGPU S. Mahmood H. Rangwala Department of Computer Science George Mason University International Conference on High Performance Computing & Communications, 2011 Ban, Canada Mahmood, Rangwala GPU-Euler
  • 2. Motivation Method Results Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 3. Motivation Genome Assembly Method Previous Work Results GPGPU Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 4. Motivation Genome Assembly Method Previous Work Results GPGPU Summary Genome Genome a biological blueprint. Very long chains of four types of nucleobases. Adenine Guanine Cytosine Thymine Important to understand the Figure: Double Helix DNA function of the organism. representation 1 1 Image courtesy of Image Library of Biological Macromolecules, Jena, Germany. http://www.imb-jena.de/IMAGE.html Mahmood, Rangwala GPU-Euler
  • 5. Motivation Genome Assembly Method Previous Work Results GPGPU Summary Sequence Assembly Challenges Total number of nucleobases in a genome is very large eg. Human Genome has 3.2 Billion base pairs. Existing technologies can only read a fraction of this long strand. Smaller fragments(reads) are required to be stitched together. Figure: Sequence Assembly 2 2 Image courtesy of Center for BioInformatics Computational Biology,UMD. www.cbcb.umd.edu/research/assembly_primer.html Mahmood, Rangwala GPU-Euler
  • 6. Motivation Genome Assembly Method Previous Work Results GPGPU Summary Problem Statement Given a set of alphabets ∑ = {A, G , C , T } and a set of strings R = {r1 , r2 , r3 . . . rn } over alphabet Σ Construct Super String S, containing all the strings from R. Similar to Shortest Common Super string. Need to consider Repeats. Massive volume of data Mahmood, Rangwala GPU-Euler
  • 7. Motivation Genome Assembly Method Previous Work Results GPGPU Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 8. Motivation Genome Assembly Method Previous Work Results GPGPU Summary Sequence Assembly Techniques Tools Greedy Assemblers VCAKE Overlap-layout-consensus Celera Eulerian Path Euler, EulerSR, Velvet ABySS Mahmood, Rangwala GPU-Euler
  • 9. Motivation Genome Assembly Method Previous Work Results GPGPU Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 10. Motivation Genome Assembly Method Previous Work Results GPGPU Summary General Purpose GPU Computing GPUs for General Purpose Computing Massive parallelism for application. nVidia CUDA , a framework for development on nVidia GPUs. Similar model for parallel computation Parallel Random Access Machine (PRAM) Figure: CUDA Application Single Instruction Multiple Data Stack (SIMD) 3 3 Image courtesy of nVidia : nVidia CUDA Toolkit Reference Manual Mahmood, Rangwala GPU-Euler
  • 11. Motivation Genome Assembly Method Previous Work Results GPGPU Summary CUDA Compute Unied Device Architecture A CUDA enabled device has Symmetric Multiprocessor (SM) Each SM has a set of Streaming Processors (SP). Global Memory. Concurrent execution of same code on all SM. Computations use GPU memory. Figure: Hardware Architecture 4 4 Image courtesy of nVidia : nVidia CUDA Toolkit Reference Manual Mahmood, Rangwala GPU-Euler
  • 12. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 13. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Concepts de-Bruijn Graph A directed graph vertice are k length word edge represents a k − 1 between vertices. Contigs Assembled sequences from the input data. EulerTour A graph traversal visiting Figure: de-Bruijn Graph each edge only once. Mahmood, Rangwala GPU-Euler
  • 14. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Objective Represent input read as a de-Bruijn graph Each edge would correspond to a single base. An Euler tour will visit each base only once. Mahmood, Rangwala GPU-Euler
  • 15. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Parallel Eulerian Assembly FASTA file (input) Reads Construct de-Bruijn Graph Debruijn Graph Construction Graph Euler Tour Construction Annotated Graph Identify Contigs Find Euler Tour EulerGPU Contigs Output Contigs FASTA file (output) Figure: GPU Euler Work ow Mahmood, Rangwala GPU-Euler
  • 16. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Parallel de-Bruijn Graph Construction FASTA file CUDA Assign each CUDA thread to Count Edges one read. CUDA Generate k -mers and Setup Vertices k + 1-mers. Store them in a hash table. CUDA Setup Create nodes from k-mers Edges and vertices from k + 1-mers. Figure: Graph Construction Mahmood, Rangwala GPU-Euler
  • 17. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Parallel Euler Tour Create a Edge Successor Graph from de-Bruijn Graph. CUDA Identify Circuits in the Edge Assign DeBruijn Graph Successor Successor Graph Annotated Graph Create a Circuit Graph by CUDA Create Circuit CUDA identifying adjacent circuits. Comp. Label Graph Find Component Circuit Graph Calculate a spanning tree for Find Circuit. Spanning Tree Spanning Tree CUDA Execute Swipe Euler Tour Traverse Circuit Graph and switch successor edges of Figure: Parallel Euler Tour adjacent Circuits. Mahmood, Rangwala GPU-Euler
  • 18. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Decomposition of Dierent Phases Phase Computation I/O and k-mer Extraction CPU + GPU Hash Table Construction GPU debruijn Graph Construction GPU Euler Tour Construction GPU + CPU Sub-steps for Euler Tour Construction Finding Connected Component GPU Circuit Graph Creation GPU Spanning Tree CPU Swipe Execution GPU Traversal (Other) GPU Contig Generation (O/P) CPU Mahmood, Rangwala GPU-Euler
  • 19. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 20. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Time Complexity Analysis Step Complexity Processors de-Bruijn Graph Construction O (1) O (n ) Euler Tour Construction (log n) O (n ) Spanning Tree O (log |V |) O (|V |) GPU-Euler O (log n ) O (n ) Mahmood, Rangwala GPU-Euler
  • 21. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 22. Motivation Parallel Eulerian Assembly Method Time Complexity Analysis Results Evaluation Summary Experimental Protocol Compared Timing, N50 Score, Mean length with EulerSR using various parameters. Why EulerSR Based on same concept Shared memory approach Support short reads Contigs with length 100 were included in the comparison. Calculated contig converge using MUMMER. Individual GPU Computations were timed as well. Mahmood, Rangwala GPU-Euler
  • 23. Motivation Method Data sets Results Results Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 24. Motivation Method Data sets Results Results Summary Data Sets Genome size and number of simulated reads for dierent read length Genome Length 36 bp 50 bp 256 bp Campylobacter Jejuni 1,641,481 911,934 656,593 128,241 Neisseria Meningitidis 2,184,406 1,213,559 873,763 170,657 Lactococcus Lactisd 2,635,589 1,314,216 946,236 184,812 Mahmood, Rangwala GPU-Euler
  • 25. Motivation Method Data sets Results Results Summary Outline 1 Motivation Genome Assembly Previous Work GPGPU 2 Method Parallel Eulerian Assembly Time Complexity Analysis Evaluation 3 Results Data sets Results Mahmood, Rangwala GPU-Euler
  • 26. Motivation Method Data sets Results Results Summary Results Execution Time Comparison Runtime Comparison for Campylobacter Jejuni Runtime Comparison for Neisseria Meningitidis 180 300 EulerSR EulerSR EulerSR* EulerSR* GPU-Euler GPU-Euler 160 250 140 Assembly Time (seconds) Assembly Time (seconds) 200 120 100 150 80 100 60 50 40 20 0 36bp 50bp 256bp 36bp 50bp 256bp Read Length Read Length Runtime Comparison for Lactococcus Lactis 300 EulerSR EulerSR* GPU-Euler 250 Assembly Time (seconds) 200 150 100 50 0 36bp 50bp 256bp Read Length Mahmood, Rangwala GPU-Euler
  • 27. Motivation Method Data sets Results Results Summary Results N50 Score Comparison N50 Score Comparison for Campylobacter Jejuni N50 Score Comparison for Neisseria Meningitidis 120000 35000 EulerSR EulerSR EulerSR* EulerSR* GPU-Euler GPU-Euler 30000 100000 25000 80000 N50 Score (bases) N50 Score (bases) 20000 60000 15000 40000 10000 20000 5000 0 0 36bp 50bp 256bp 36bp 50bp 256bp Read Length Read Length N50 Score Comparison for Lactococcus Lactis 80000 EulerSR EulerSR* GPU-Euler 70000 60000 N50 Score (bases) 50000 40000 30000 20000 10000 0 36bp 50bp 256bp Read Length Mahmood, Rangwala GPU-Euler
  • 28. Motivation Method Data sets Results Results Summary Results Accuracy Comparison Campylobacter Jejuni Neisseria Meningitidis 100 100 EulerSR EulerSR EulerSR* EulerSR* GPU-Euler 90 GPU-Euler 95 80 70 Weighted Accuracy Weighted Accuracy 90 60 50 85 40 30 80 20 10 75 0 36bp 50bp 256bp 36bp 50bp 256bp Read Length Read Length Lactococcus Lactis 100 EulerSR EulerSR* GPU-Euler 95 Weighted Accuracy 90 85 80 75 36bp 50bp 256bp Read Length Mahmood, Rangwala GPU-Euler
  • 29. Motivation Method Data sets Results Results Summary GPU Euler Phase Distribution Phase Computation % Time I/O and k-mer Extraction CPU + GPU 77.29+1.44 Hash Table Construction GPU 0.31 debruijn Graph Construction GPU 1.15 Euler Tour Construction GPU + CPU Sub-steps for Euler Tour Construction Finding Connected Component GPU 10.06 Spanning Tree CPU 0.06 Swipe Execution GPU 0.01 Circuit Graph Traversal (Other) GPU 0.72 Contig Generation (O/P) CPU 4.39 Mahmood, Rangwala GPU-Euler
  • 30. Motivation Method Results Summary Summary Exploiting GPUs for Sequence Assembly. Implementation of PRAM algorithm on CUDA devices. Outlook No Error Correction Graph Simplication. Mahmood, Rangwala GPU-Euler
  • 31. Motivation Method Results Summary Questions Questions? Mahmood, Rangwala GPU-Euler
  • 32. Motivation Method Results Summary Thank you Thank you!! Mahmood, Rangwala GPU-Euler