Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Apollo Collaborative genome annotation editing


Published on

Apollo is a web-based application that supports and enables collaborative genome curation in real time, allowing teams of curators to improve on existing automated gene models through an intuitive interface. Apollo allows researchers to break down large amounts of data into manageable portions to mobilize groups of researchers with shared interests.

A Workshop at the Stowers Institute for Medical Research.

Published in: Science
  • Be the first to comment

Apollo Collaborative genome annotation editing

  1. 1. Apollo Collaborative genome annotation editing A workshop for the Stowers Institute Research Community Monica Munoz-Torres, PhD | @monimunozto Berkeley Bioinformatics Open-Source Projects (BBOP) Environmental Genomics & Systems Biology Division Lawrence Berkeley National Laboratory Kansas City, MO | 12 December, 2016
  2. 2. Outline • Today we will discuss effective ways to extract valuable information about a genome through curation efforts.
  3. 3. After this talk you will... • Better understand curation in the context of genome annotation: assembled genome à automated annotation à manual annotation • Become familiar with Apollo’s environment and functionality. • Learn to identify homologs of known genes of interest in your newly sequenced genome. • Learn how to corroborate and modify automatically annotated gene models using all available evidence in Apollo.
  4. 4. Experimental design, sampling. Comparative analyses Merged Gene Set Manual Annotation Automated Annotation Sequencing Assembly Synthesis & dissemination. Genome sequencing projects
  5. 5. Unlocking genomes Marbach et al. 2011. Nature Methods | | Alexander Wild
  6. 6. First, a refresher
  7. 7. A few things to remember during curation of gene models 7BIO-REFRESHER • KEEP A GLOSSARY HANDY from contig to splice site • WHAT IS A GENE? defining your goal • TRANSCRIPTION mRNA in detail • TRANSLATION reading frames, etc. • GENOME CURATION steps involved
  8. 8. The gene: a moving target “The gene is a union of genomic sequences encoding a coherent set of potentially overlapping functional products.” Gerstein et al., 2007. Genome Res
  9. 9. 9 "Gene structure" by Daycd- Wikimedia Commons BIO-REFRESHER mRNA • Although of brief existence, understanding mRNAs is crucial, as they will become the center of your work.
  10. 10. 10BIO-REFRESHER Reading frames v In eukaryotes, only one reading frame per section of DNA is biologically relevant at a time: it has the potential to be transcribed into RNA and translated into protein. This is called the OPEN READING FRAME (ORF) • ORF = Start signal + coding sequence (divisible by 3) + Stop signal
  11. 11. 11BIO-REFRESHER Splice sites v The spliceosome catalyzes the removal of introns and the ligation of flanking exons. v Splicing signals (from the point of view of an intron): • One splice signal (site) on the 5’ end: usually GT (less common: GC) • And a 3’ end splice site: usually AG • Canonical splice sites look like this: …]5’-GT/AG-3’[…
  12. 12. 12BIO-REFRESHER Exons and Introns v Introns can interrupt the reading frame of a gene by inserting a sequence between two consecutive codons v Between the first and second nucleotide of a codon v Or between the second and third nucleotide of a codon "Exon and Intron classes”. Licensed under Fair use via Wikipedia
  13. 13. 13BIO-REFRESHER Obstacles to transcription and translation v The presence of premature Stop codons in the message is possible. A process called non-sense mediated decay checks for them and corrects them to avoid: incomplete splicing, DNA mutations, transcription errors, and leaky scanning of ribosome – causing changes in the reading frame (frame shifts). v Insertions and deletions (indels) can cause frame shifts when indel is not divisible by three. As a result, the peptide can be abnormally long, or abnormally short – depending when the first in-frame Stop signal is located.
  14. 14. Prediction & Annotation
  15. 15. 15GENE PREDICTION & ANNOTATION PREDICTION & ANNOTATION v Identification and annotation of genome features: • primarily focuses on protein-coding genes. • also identifies RNAs (tRNA, rRNA, long and small non-coding RNAs (ncRNA)), regulatory motifs, repetitive elements, etc. • happens in 2 phases: 1. Computation phase 2. Annotation phase
  16. 16. 16GENE PREDICTION & ANNOTATION COMPUTATION PHASE a. Experimental data are aligned to the genome: expressed sequence tags, RNA-sequencing reads, proteins (also from other species). b. Gene predictions are generated: - ab initio: based on nucleotide sequence and composition e.g. Augustus, GENSCAN, geneid, fgenesh, etc. - evidence-driven: identifying also domains and motifs e.g. SGP2, JAMg, fgenesh++, etc. Result: the single most likely coding sequence, no UTRs, no isoforms. Yandell & Ence. Nature Rev 2012 doi:10.1038/nrg3174
  17. 17. 17GENE PREDICTION & ANNOTATION ANNOTATION PHASE Experimental data (evidence) and predictions are synthetized into gene annotations. Result: gene models that generally include UTRs, isoforms, evidence trails. Yandell & Ence. Nature Rev 2012 doi:10.1038/nrg3174 5’ UTR 3’ UTR
  18. 18. 18 In some cases algorithms and metrics used to generate consensus sets may actually reduce the accuracy of the gene’s representation. CONSENSUS GENE SETS Gene models may be organized into sets using: v combiners for automatic integration of predicted sets e.g: GLEAN, EvidenceModeler or v tools packaged into pipelines e.g: MAKER, PASA, Gnomon, Ensembl, etc. GENE PREDICTION & ANNOTATION
  19. 19. 19BIO-REFRESHER Good genes are required! 1. Generate gene models v A few rounds of gene prediction. 2. Annotate gene models v Function, expression patterns, metabolic network memberships. 3. Manually review them v Structure & Function.
  20. 20. Best representation of biology & removal of elements reflecting errors in automated analyses. Functional assignments through comparative analysis using literature, databases, and experimental data. Apollo Gene Ontology Curation improves quality
  21. 21. 21BIO-REFRESHER Curation is inherently collaborative • It is impossible for a single individual to curate an entire genome with precise biological fidelity. • Curators need second opinions and insights from colleagues with domain and gene family expertise.
  22. 22. Collaborative curation with Apollo
  23. 23. Apollo: genome annotation editing • Collaborative, instantaneous, web-based, built on top of JBrowse. • Supports real time collaboration & generates analysis-ready data USER-CREATED ANNOTATIONS EVIDENCE TRACKS ANNOTATOR PANEL
  24. 24. BECOMING ACQUAINTED WITH APOLLO General process of curation 1. Select or find a region of interest (e.g. scaffold). 2. Select appropriate evidence tracks to review the genome element to annotate (e.g. gene model). 3. Determine whether a feature in an existing evidence track will provide a reasonable gene model to start working. 4. If necessary, adjust the gene model. 5. Check your edited gene model for integrity and accuracy by comparing it with available homologs. 6. Comment and finish.
  25. 25. Color by CDS frame, toggle strands, set color scheme and highlights. Upload evidence files (GFF3, BAM, BigWig), add combination and sequence search tracks. Query the genome using BLAT. Navigation and zoom. Search for a gene model or a scaffold. User-created annotations. Annotator panel. Evidence Tracks. Stage and cell-type specific transcription data. Apollo Genome Annotation Editor Protein coding, pseudogenes, ncRNAs, regulatory elements, variants, etc. Admin.
  26. 26. Apollo Architecture Web-based client + annotation-editing engine + server-side data service
  27. 27. Generates ready-made computable data Ref Sequence
  28. 28. Supports real time collaboration
  29. 29. Let’s Play!
  30. 30. Access Apollo
  31. 31. Annotations Organism Users Groups AdminTracks Reference Sequence Removable side dock
  32. 32. 1 Annotations gene mRNA Annotation details & exon boundaries 1 2 2
  33. 33. Curating with Apollo
  34. 34. 34 | BECOMING ACQUAINTED WITH APOLLO USER NAVIGATION Annotator panel. • Choose appropriate evidence from list of “Tracks” on annotator panel. • Select & drag elements from evidence track into the ‘User-created Annotations’ area. • Hovering over annotation in progress brings up an information pop-up. • Creating a new annotation
  35. 35. Adding a gene model
  36. 36. Adding a gene model
  37. 37. Adding a gene model
  38. 38. Editing functionality
  39. 39. Editing functionality Example: Adding an exon supported by experimental data • RNAseq reads show evidence in support of a transcribed product that was not predicted. • Add exon by dragging up one of the RNAseq reads.
  40. 40. Editing functionality Example: Adjusting exon boundaries supported by experimental data
  41. 41. 41 | USER NAVIGATION BECOMING ACQUAINTED WITH APOLLO • ‘Zoom to base level’ reveals the DNA Track.
  42. 42. 42 | USER NAVIGATION BECOMING ACQUAINTED WITH APOLLO • Color exons by CDS from the ‘View’ menu.
  43. 43. 43 | Zoom in/out with keyboard: shift + arrow keys up/down USER NAVIGATION BECOMING ACQUAINTED WITH APOLLO • Toggle reference DNA sequence and translation frames in forward strand. Toggle models in either direction.
  44. 44. annotating simple cases
  45. 45. “Simple case”: - the predicted gene model is correct or nearly correct, and - this model is supported by evidence that completely or mostly agrees with the prediction. - evidence that extends beyond the predicted model is assumed to be non-coding sequence. The following are simple modifications. ANNOTATING SIMPLE CASES BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  46. 46. • A confirmation box will warn you if the receiving transcript is not on the same strand as the feature where the new exon originated. • Check ‘Start’ and ‘Stop’ signals after each edit. ADDING EXONS BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  47. 47. If transcript alignment data are available & extend beyond your original annotation, you may extend or add UTRs. 1. Right click at the exon edge and ‘Zoom to base level’. 2. Place the cursor over the edge of the exon until it becomes a black arrow then click and drag the edge of the exon to the new coordinate position that includes the UTR. ADDING UTRs To add a new spliced UTR to an existing annotation also follow the procedure for adding an exon. BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  48. 48. To modify an exon boundary and match data in the evidence tracks: select both the [offending] exon and the feature with the expected boundary, then right click on the annotation to select ‘Set 3’ end’ or ‘Set 5’ end’ as appropriate. In some cases all the data may disagree with the annotation, in other cases some data support the annotation and some of the data support one or more alternative transcripts. Try to annotate as many alternative transcripts as are well supported by the data. MATCHING EXON BOUNDARY TO EVIDENCE BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  49. 49. 1. Two exons from different tracks sharing the same start/end coordinates display a red bar to indicate matching edges. 2. Selecting the whole annotation or one exon at a time, use this edge- matching function and scroll along the length of the annotation, verifying exon boundaries against available data. Use square [ ] brackets to scroll from exon to exon. User curly { } brackets to scroll from annotation to annotation. 3. Check if cDNA / RNAseq reads lack one or more of the annotated exons or include additional exons. CHECKING EXON INTEGRITY BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  50. 50. Non-canonical splice sites flags. Double click: selection of feature and sub-features Evidence Tracks Area ‘User-created Annotations’ Track Edge-matching Apollo’s editing logic (brain): § selects longest ORF as CDS § flags non-canonical splice sites ORFs AND SPLICE SITES BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  51. 51. Non-canonical splices are indicated by an orange circle with a white exclamation point inside, placed over the edge of the offending exon. Canonical splice sites: 3’-…exon]GA / TG[exon…-5’ 5’-…exon]GT / AG[exon…-3’ reverse strand, not reverse-complemented: forward strand SPLICE SITES Zoom to review non-canonical splice site warnings. Although these may not always have to be corrected (e.g GC donor), they should be flagged with a comment. Exon/intron splice site error warning Curated model BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  52. 52. Apollo calculates the longest possible open reading frame (ORF) that includes canonical ‘Start’ and ‘Stop’ signals within the predicted exons. If ‘Start’ appears to be incorrect, modify it by selecting an in-frame ‘Start’ codon further up or downstream, depending on evidence (proteins, RNAseq). It may be present outside the predicted gene model, within a region supported by another evidence track. In very rare cases, the actual ‘Start’ codon may be non-canonical (non-ATG). ‘Start’ AND ‘Stop’ SITES BECOMING ACQUAINTED WITH APOLLO SIMPLE CASES
  53. 53. annotating complex cases
  54. 54. Evidence may support joining two or more different gene models. Warning: protein alignments may have incorrect splice sites and lack non-conserved regions! 1. In ‘User-created Annotations’ area shift-click to select an intron from each gene model and right click to select the ‘Merge’ option from the menu. 2. Drag supporting evidence tracks over the candidate models to corroborate overlap, or review edge matching and coverage across models. 3. Check the resulting translation by querying a protein database e.g. UniProt, NCBI nr. Add comments to record that this annotation is the result of a merge. Red lines around exons: ‘edge-matching’ allows annotators to confirm whether the evidence is in agreement without examining each exon at the base level. COMPLEX CASES merge two gene predictions on the same scaffold BECOMING ACQUAINTED WITH APOLLO COMPLEX CASES
  55. 55. One or more splits may be recommended when: - different segments of the predicted protein align to two or more different gene families - predicted protein doesn’t align to known proteins over its entire length - Transcript data may support a split, but first verify whether they are alternative transcripts. COMPLEX CASES split a gene prediction BECOMING ACQUAINTED WITH APOLLO COMPLEX CASES
  56. 56. DNA Track ‘User-created Annotations’ Track COMPLEX CASES annotate frameshifts and correct single-base errors Always remember: when annotating gene models using Apollo, you are looking at a ‘frozen’ version of the genome assembly and you will not be able to modify the assembly itself. BECOMING ACQUAINTED WITH APOLLO COMPLEX CASES
  57. 57. COMPLEX CASES correcting selenocysteine containing proteins BECOMING ACQUAINTED WITH APOLLO COMPLEX CASES
  58. 58. COMPLEX CASES correcting selenocysteine containing proteins BECOMING ACQUAINTED WITH APOLLO COMPLEX CASES
  59. 59. 1. Apollo allows annotators to make single base modifications or frameshifts that are reflected in the sequence and structure of any transcripts overlapping the modification. These manipulations do NOT change the underlying genomic sequence. 2. If you determine that you need to make one of these changes, zoom in to the nucleotide level and right click over a single nucleotide on the genomic sequence to access a menu that provides options for creating insertions, deletions or substitutions. 3. The ‘Create Genomic Insertion’ feature will require you to enter the necessary string of nucleotide residues that will be inserted to the right of the cursor’s current location. The ‘Create Genomic Deletion’ option will require you to enter the length of the deletion, starting with the nucleotide where the cursor is positioned. The ‘Create Genomic Substitution’ feature asks for the string of nucleotide residues that will replace the ones on the DNA track. 4. Once you have entered the modifications, Apollo will recalculate the corrected transcript and protein sequences, which will appear when you use the right-click menu ‘Get Sequence’ option. Since the underlying genomic sequence is reflected in all annotations that include the modified region you should alert the curators of your organisms database using the ‘Comments’ section to report the CDS edits. 5. In special cases such as selenocysteine containing proteins (read-throughs), right-click over the offending/premature ‘Stop’ signal and choose the ‘Set readthrough stop codon’ option from the menu. COMPLEX CASES annotating frameshifts and correcting single-base errors & selenocysteines BECOMING ACQUAINTED WITH APOLLO COMPLEX CASES
  62. 62. The Annotation Information Editor • Add PubMed IDs • Include GO terms as appropriate from any of the three ontologies • Write comments stating how you have validated each model. USER NAVIGATION BECOMING ACQUAINTED WITH APOLLO
  63. 63. 63 | USER NAVIGATION BECOMING ACQUAINTED WITH APOLLO • Keeping track of each edit
  64. 64. Annotations, annotation edits, and History: stored in a centralized database. USER NAVIGATION BECOMING ACQUAINTED WITH APOLLO
  65. 65. Follow the checklist until you are happy with the annotation! And remember to… – comment to validate your annotation, even if you made no changes to an existing model. Think of comments as your vote of confidence. – or add a comment to inform the community of unresolved issues you think this model may have. 65 | Always Remember: Apollo curation is a community effort so please use comments to communicate the reasons for your annotation. Your comments will be visible to everyone. COMPLETING THE ANNOTATION BECOMING ACQUAINTED WITH APOLLO
  66. 66. Checklist
  67. 67. • Check ‘Start’ and ‘Stop’ sites. • Check splice sites: most splice sites display these residues …]5’-GT/AG-3’[… • Check if you can annotate UTRs, for example using RNA-Seq data: – align it against relevant genes/gene family – blastp against NCBI’s RefSeq or nr • Check for gaps in the genome. • Additional functionality may be necessary: – merging 2 gene predictions - same scaffold – ‘merging’ 2 gene predictions - different scaffolds – splitting a gene prediction – annotating frameshifts – annotating selenocysteines, correcting single-base and other assembly errors, etc. 67 | • Add: – Important project information in the form of comments – IDs from public databases e.g. GenBank (via DBXRef), gene symbol(s), common name(s), synonyms, top BLAST hits, orthologs with species names, and everything else you can think of, because you are the expert. – Comments about the kinds of changes you made to the gene model of interest, if any. – Any appropriate functional assignments, e.g. via BLAST, RNA-Seq data, literature searches, etc. CHECKLIST for accuracy and integrity MANUAL ANNOTATION CHECKLIST
  68. 68. Example
  69. 69. What’s new?... finding inspiration in the literature. Example 69 “Molecular analysis of bed bug populations from across the USA and Europe found that >80% and >95% of the respective populations contained V419L and/or L925I mutations in the voltage-gated sodium channel gene, indicating widespread distribution of target-site-based pyrethroid resistance.”
  70. 70. Example Example 70 Curation example using the Hyalella azteca genome (amphipod crustacean).
  71. 71. What we know about this genome • Data at NCBI: • >37,000 nucleotide seqsà scaffolds, mitochondrial genes • 344 amino acid seqsà mitochondrion • 47 ESTs • 0 conserved domains identified • 0 “gene” entries submitted • Data at i5K Workspace@NAL (annotation hosted at USDA) - 10,832 scaffolds: 23,288 transcripts: 12,906 proteins Example 71
  72. 72. PubMed Search: what’s new? Example 72
  73. 73. PubMed Search: what’s new? Example 73 “Ten populations differed by at least 550-fold in sensitivity to pyrethroids.” “Sequencing the primary pyrethroid target site, the voltage- gated sodium channel (vgsc), shows that point mutations and their spread in natural populations were responsible for differences in pyrethroid sensitivity.” “The finding that a non-target aquatic species has acquired resistance to pesticides used only on terrestrial pests is troubling evidence of the impact of chronic pesticide transport from land-based applications into aquatic systems.”
  74. 74. How many sequences are there, publicly available, for our gene of interest? Example 74 • Para, (voltage-gated sodium channel alpha subunit; Nasonia vitripennis). • NaCP60E (Sodium channel protein 60 E; D. melanogaster). – MF: voltage-gated cation channel activity (IDA, GO:0022843). – BP: olfactory behavior (IMP, GO:0042048), sodium ion transmembrane transport (ISS,GO:0035725). – CC: voltage-gated sodium channel complex (IEA, GO:0001518). And what do we know about them?
  77. 77. BLAT search results Example 77 • High-scoring segment pairs (hsp) are listed in tabulated format. • Clicking on one line of results sends you to those coordinates.
  78. 78. Creating a new gene model: drag and drop Example 78 • Apollo automatically calculates longest ORF. • In this case, ORF includes the high-scoring segment pairs (hsp), marked here in blue. • Note that gene is transcribed from reverse strand.
  79. 79. Available evidence tracks Example 79
  80. 80. Get Sequence Example 80
  81. 81. Also, flanking sequences (other gene models) vs. NCBI nr Example 81 In this case, two gene models upstream, at 5’ end. BLAST hsps
  82. 82. Review alignments Example 82 HaztTmpM006234 HaztTmpM006233 HaztTmpM006232
  83. 83. Hypothesis for vgsc gene model Example 83
  84. 84. Editing: merge the three models Example 84 Merge by dropping an exon or gene model onto another. Merge by selecting two exons (holding down “Shift”) and using the right click menu. or…
  85. 85. Result of merging the gene models: Example 85
  86. 86. Editing: correct offending splice site Example 86 Modify exon / intron boundary: - Drag the end of the exon to the nearest canonical splice site. or - Use right-click menu.
  87. 87. Editing: set translation start Example 87
  88. 88. Editing: delete exon not supported by evidence Example 88 Delete first exon from HaztTmpM006233
  89. 89. Editing: add an exon supported by RNAseq Example 89 • RNAseq reads show evidence in support of transcribed product, which was not predicted. • Add exon at coordinates 97946-98012 by dragging up one of the RNAseq reads.
  90. 90. Editing: adjust offending splice site using evidence Example 90
  91. 91. Editing: adjust other boundaries supported by evidence Example 91
  92. 92. Finished model Example 92 Corroborate integrity and accuracy of the model: - Start and Stop - Exon structure and splice sites …]5’-GT/AG-3’[… - Check the predicted protein product vs. NCBI nr, UniProt, etc.
  93. 93. Information Editor • DBXRefs: e.g. NP_001128389.1, N. vitripennis, RefSeq • PubMed identifier: PMID: 24065824 • Gene Ontology IDs: GO:0022843, GO:0042048, GO:0035725, GO:0001518. • Comments • Name, Symbol • Approve / Delete radio button Example 93 Comments (if applicable)
  94. 94. Exercises
  95. 95. Example dataset: Apis mellifera
  96. 96. Practice using the Apis mellifera genome 1. Evidence in support of protein coding gene models. 1.1 Consensus Gene Sets: Official Gene Set v3.2 Official Gene Set v1.0 1.2 Consensus Gene Sets comparison: OGSv3.2 genes that merge OGSv1.0 and RefSeq genes OGSv3.2 genes that split OGSv1.0 and RefSeq genes 1.3 Protein Coding Gene Predictions Supported by Biological Evidence: NCBI Gnomon Fgenesh++ with RNASeq training data Fgenesh++ without RNASeq training data NCBI RefSeq Protein Coding Genes and Low Quality Protein Coding Genes 1.4 Ab initio protein coding gene predictions: Augustus Set 12, Augustus Set 9, Fgenesh, GeneID, N-SCAN, SGP2 1.5 Transcript Sequence Alignment: NCBI ESTs, Apis cerana RNA-Seq, Forager Bee Brain Illumina Contigs, Nurse Bee Brain Illumina Contigs, Forager RNA-Seq reads, Nurse RNA-Seq reads, Abdomen 454 Contigs, Brain and Ovary 454 Contigs, Embryo 454 Contigs, Larvae 454 Contigs, Mixed Antennae 454 Contigs, Ovary 454 Contigs, Testes 454 Contigs, Forager RNA-Seq HeatMap, Forager RNA-Seq XY Plot, Nurse RNA- Seq HeatMap, Nurse RNA-Seq XY Plot
  97. 97. Practice using the Apis mellifera genome 1. Evidence in support of protein coding gene models (Continued). 1.6 Protein homolog alignment: Acep_OGSv1.2 Aech_OGSv3.8 Cflo_OGSv3.3 Dmel_r5.42 Hsal_OGSv3.3 Lhum_OGSv1.2 Nvit_OGSv1.2 Nvit_OGSv2.0 Pbar_OGSv1.2 Sinv_OGSv2.2.3 Znev_OGSv2.1 Metazoa_Swissprot 2. Evidence in support of non protein coding gene models 2.1 Non-protein coding gene predictions: NCBI RefSeq Noncoding RNA NCBI RefSeq miRNA 2.2 Pseudogene predictions: NCBI RefSeq Pseudogene
  98. 98. Apollo Development Nathan Dunn Technical Lead Eric Yao Christine Elsik’s Lab, University of Missouri Suzi Lewis Principal Investigator BBOP Moni Munoz-Torres Project Manager Deepak Unni JBrowse. Ian Holmes’ Lab University of California, Berkeley
  99. 99. Thank you! Berkeley Bioinformatics Open-Source Projects, Environmental Genomics & Systems Biology, Lawrence Berkeley National Laboratory Funding • Apollo is supported by NIH grants 5R01GM080203 from NIGMS, and 5R01HG004483 from NHGRI. • BBOP is also supported by the Director, Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231 Collaborators • Ian Holmes, Eric Yao, UC Berkeley (JBrowse) • Chris Elsik, Deepak Unni, U of Missouri (Apollo) • Monica Poelchau, USDA/NAL (Apollo) • i5k Community UNIVERSITY OF CALIFORNIA Suzanna Lewis & Chris Mungall Seth Carbon (Noctua / AmiGO) Nathan Dunn (Apollo) Monica Munoz-Torres (Apollo / GO) Jeremy Nguyen Xuan (Monarch Init.)
  100. 100. PUBLIC DEMO 100 | APOLLO ON THE WEB instructions Public Honey bee demo available at: Username: Password: demo
  101. 101. APOLLO demonstration PUBLIC DEMO 101 Demonstration video available at
  102. 102. Access Apollo