Next	
  Genera*on	
  Sequencing:	
  
Reference	
  Material	
  (RM)	
  Selec*on	
  and	
  
                 Design	
  

                     Workgroup	
  
   Genome	
  in	
  a	
  BoAle	
  Consor*um	
  /	
  NIST	
  
                  August	
  16-­‐17,	
  2012	
  
RM	
  Scope	
  
•  Human	
  Genome	
  &	
  Tumor	
  Sequencing	
  

•  Variant	
  Types	
  
   –  SNP	
  
   –  InDel	
  /	
  Subs*tu*on	
  
   –  CNV	
  
   –  Structural	
  variant	
  


                                                     | 2
Reference	
  Material	
  Needed	
  For	
  
•  PlaVorm	
  valida*on	
  
    –  Sequencing	
  System	
  
    –  Bioinforma*cs/Analysis	
  Pipeline	
  

•  Test	
  valida*on	
  
    –  Whole	
  genome	
  
    –  Targeted	
  
    –  Germline	
  vs.	
  tumor	
  

•  Reagent	
  valida*on	
  

                                                 | 3
New	
  York	
  State	
  proposed	
  dra[	
  
guidelines	
  for	
  NGS	
  clinical	
  cancer	
  tests	
  	
  
•  …	
  A	
  posi%ve/sensi%vity	
  control	
  should	
  be	
  included	
  to	
  
   verify	
  analy*c	
  sensi*vity.	
  We	
  suggest	
  including	
  an	
  
   individually	
  bar	
  coded	
  low	
  posi*ve	
  control	
  near	
  the	
  
   sensi*vity	
  of	
  the	
  assay,	
  containing	
  mul*ple	
  known	
  soma*c	
  
   altera*ons	
  of	
  each	
  kind	
  to	
  be	
  detected,	
  to	
  verify	
  that	
  low	
  
   percentage	
  gene*c	
  variants	
  can	
  be	
  detected	
  in	
  every	
  run.	
  …	
  
•  Performance	
  characteris*cs	
  for	
  each	
  type	
  of	
  gene*c	
  
   altera*on	
  the	
  assay	
  is	
  intended	
  to	
  detect,	
  e.g.	
  SNVs,	
  indels,	
  
   CNVs,	
  must	
  be	
  established	
  and	
  validated	
  
•  Sequence	
  a	
  well-­‐characterized	
  reference	
  sample	
  (e.g.	
  
   HapMap	
  DNA	
  GM12878)	
  to	
  determine	
  error	
  rate	
  and	
  
   establish	
  depth/uniformity	
  of	
  coverage	
  across	
  all	
  
   amplicons.	
  

                                                                                           | 4
Associa*on	
  for	
  Molecular	
  Pathology	
  
    Comments	
  to	
  FDA	
  UHT-­‐Sequencing	
  Mee*ng,	
  June	
  2011	
  

•  …	
  Performance	
  of	
  and	
  coverage	
  needs	
  for	
  a	
  given	
  plaVorm	
  
   are	
  likely	
  to	
  differ	
  depending	
  on	
  the	
  nucleic	
  acid	
  and	
  DNA	
  
   regions	
  analyzed,	
  the	
  variants	
  interrogated,	
  the	
  rela*ve	
  
   allele	
  propor*ons	
  of	
  par*cular	
  variants,	
  …	
  Evalua*on	
  
   should	
  consider	
  the	
  effects	
  of	
  rela*ve	
  GC	
  content,	
  
   homopolymeric	
  and	
  other	
  regions	
  of	
  repe**ve	
  sequence,	
  
   homologous	
  gene	
  regions	
  and	
  DNA	
  structural	
  variants,	
  …	
  
   This	
  necessitates	
  flexibility	
  and	
  individualiza*on	
  in	
  the	
  
   development	
  of	
  valida*on	
  protocols,	
  guidelines,	
  and	
  
   controls	
  on	
  a	
  (clinical)	
  applica*on-­‐by-­‐applica*on	
  basis.	
  …	
  	
  
•  Assay	
  controls	
  should	
  include	
  a	
  range	
  of	
  variants,	
  …	
  
   Process	
  controls	
  like	
  NA12876	
  …	
  and	
  the	
  synthe*c	
  ERCC	
  
   RNA	
  transcripts	
  from	
  NIST	
  are	
  examples	
  of	
  poten%al	
  
   standard	
  reference	
  materials.	
  …	
  

                                                                                            | 5
Reference	
  Material	
  PorVolio	
  
Whole	
  Human	
  Genome	
  
•  What	
  sources	
  of	
  RMs	
  to	
  consider	
  
      –  Available	
  for	
  research	
  and	
  for	
  profit	
  
      –  Primary/	
  cell	
  line	
  
•  What	
  extent	
  of	
  prior	
  characteriza*on	
  
•  Which	
  ethnici*es,	
  genders	
  
•  Which	
  muta*ons	
  need	
  to	
  be	
  present	
  
      –  Is	
  medical	
  relevance	
  necessary	
  
•  Ini*ally	
  to	
  have	
  
      –  ONE	
  characterized	
  genome	
  RM	
  	
  	
  -­‐	
  or	
  
      –  Mul*ple	
  genomes,	
  lower	
  level	
  of	
  characteriza*on	
  
•  Source	
  of	
  commercial	
  development	
  and	
  distribu*on	
  
      –  Manufactured	
  under	
  quality	
  system	
  for	
  diagnos*c	
  applica*ons	
  


                                                                                             | 6
Workgroup	
  AAendees	
  
•  Approximately	
  25	
  aAendees	
  
   –  Federal,	
  incl.	
  FDA,	
  CDA,	
  NIST	
  
   –  Lab	
  accredita*on	
  
   –  Clinical	
  reference	
  labs	
  
   –  PlaVorm	
  technologies	
  
   –  Reference	
  material	
  /	
  reagent	
  providers	
  
   –  Research	
  



                                                               | 7
Workgroup	
  Summary	
  (1)	
  
                Reference	
  Material	
  Use	
  
•  Characterize	
  PlaVorms	
  &	
  Methods	
  	
  
    –  DNA	
  sequencing	
  
    –  Exis*ng	
  &	
  upcoming	
  NGS	
  technologies	
  
    –  Research	
  applica*ons	
  
    –  PlaVorm	
  &	
  methods	
  characteriza*on	
  for	
  clinical	
  
       diagnos*cs	
  applica*ons	
  

•  Not	
  intended	
  as	
  reference	
  material	
  for	
  
    –  Valida*on	
  of	
  specific	
  muta*ons	
  in	
  a	
  panels	
  

                          Confidential – Do not copy or distribute         | 8
Workgroup	
  Summary	
  (2)	
  
                  Selected	
  Sample	
  
•  NA12878	
  	
  
    –  6,000	
  controlled	
  aliquots	
  of	
  10ug	
  each	
  on	
  order	
  by	
  
       NIST	
  from	
  Coriell	
  




                                                                                 | 9
Workgroup	
  Summary	
  (2)	
  
                  Selected	
  Sample	
  
•  NA12878	
  	
  
    –  6,000	
  controlled	
  aliquots	
  of	
  10ug	
  each	
  on	
  order	
  by	
  
       NIST	
  from	
  Coriell	
  




HOWEVER ….                                                                       | 10
Workgroup	
  Summary	
  (3)	
  
                                    Consent	
  
•  Consent	
  available	
  for	
  	
  
    –  Research	
  	
  ü	
  
•  Consent	
  not	
  available	
  for	
  
    –  Commercial	
  
         •  Incl.	
  altera*ons,	
  re-­‐distribu*on	
  
    –  Re-­‐iden*fica*on	
  through	
  sequence	
  data	
  

  NA12878 and family requires re-consenting for
  commercial use and potential re-identification
           through sequence data
                                                             | 11
Workgroup	
  Summary	
  (4)	
  
                Sample	
  &	
  Consent	
  
•  If	
  re-­‐consen*ng	
  of	
  NA12878	
  not	
  possible,	
  
   replace	
  with	
  other	
  sample	
  (Trio)	
  

•  If	
  necessary,	
  do	
  sample	
  change	
  now,	
  since	
  
   early	
  in	
  process	
  of	
  selec*ng	
  reference	
  sample	
  
•  New	
  characteriza*on	
  of	
  reference	
  material	
  
   will	
  be	
  in	
  much	
  greater	
  depth	
  and	
  with	
  
   technologies	
  that	
  are	
  more	
  advanced	
  than	
  
   exis*ng	
  data	
  for	
  NA12878	
  
                                                                    | 12
Workgroup	
  Summary	
  (5)	
  
               Sample	
  Characteris*cs	
  
•  Sample	
  characteris*cs	
  are	
  more	
  important	
  
   than	
  selec*on	
  of	
  specific	
  sample	
  IDs	
  
•  More	
  reference	
  samples	
  preferred	
  over	
  fewer	
  
   samples	
  
   –  Prefer	
  8	
  fully	
  characterized	
  samples	
  at	
  high	
  depth	
  
      and	
  corresponding	
  trios	
  at	
  lower	
  depth	
  over	
  4	
  
      fully	
  characterized	
  samples	
  plus	
  trios	
  



                                                                              | 13
Workgroup	
  Summary	
  (5)	
  
                       Sample	
  Characteris*cs	
  (cont.)	
  
•  High	
  Priority	
  
      –  Mul*ple	
  ethnici*es	
  
             •  Diversity	
  in	
  structural	
  varia*on	
  to	
  stress	
  systems	
  
      –  Balanced	
  female	
  to	
  male	
  ra*o	
  
      –  Cell	
  line,	
  low	
  passages	
  
             •  Replenish	
  supply	
  

•  Nice	
  to	
  have	
  
      –  Interracial	
  marriage	
  samples	
  	
  
             •  Controlled	
  admixture	
  


•  Less	
  cri*cal	
  
      –  Phenotypic	
  characteriza*on	
  
             •  Reference	
  material	
  not	
  for	
  discovery	
  
      –  Access	
  to	
  RNA	
  or	
  *ssues	
  
             •  No	
  limitless	
  supply	
  of	
  material	
  with	
  iden*cal	
  characteris*cs	
  



                                                                                                        | 14
Workgroup	
  Summary	
  (6)	
  
                   DNA	
  RM	
  Requirements	
  
•  DNA	
  from	
  low	
  passage	
  cell	
  lines	
  
     –  New	
  batches	
  from	
  low	
  passage	
  aliquots	
  
•  Modify	
  DNA	
  purifica*on	
  in	
  future	
  to	
  keep	
  step	
  with	
  new	
  NGS	
  
   technologies	
  
     –  Current	
  purified	
  DNA	
  fragment	
  sizes	
  are	
  80-­‐100kb	
  
          •  OK	
  for	
  exis*ng	
  technologies	
  
     –  New	
  nanopore	
  technologies	
  may	
  need	
  Mbp	
  fragments	
  
          •  Agarose	
  embedding	
  is	
  proven	
  extrac*on	
  technology	
  
•  Consider	
  footprint	
  analysis	
  of	
  all	
  batches	
  prior	
  to	
  distribu*on	
  
     –  Iden*fy	
  gene*c	
  dri[,	
  mix	
  ups,	
  ….	
  ,	
  develop	
  benchmarks	
  
•  Reference	
  material	
  that	
  mimics	
  tumor	
  sample	
  characteris*cs	
  
     –  FFPE	
  embedded	
  cells?	
  


                                                                                            | 15
Workgroup	
  Summary	
  (7)	
  
                         Sample	
  Source	
  Sugges*ons	
  
Most	
  support	
  
•  Personal	
  Genome	
  Project	
  Samples	
  	
  
      –    Includes	
  trios	
  	
  
      –    Use	
  sequence	
  data	
  to	
  derive	
  admixture	
  
      –    hAp://www.personalgenomes.org	
  
      –    Consent	
  includes	
  research	
  &	
  commercial	
  use	
  and	
  re-­‐iden*fica*on	
  
	
  
Some	
  support	
  (if	
  consent	
  sufficient)	
  
•  HS1011	
  
      –  Charcot	
  Marie	
  Tooth	
  cell	
  line	
  
             •  Lupski	
  et	
  al,	
  NEJM	
  2010	
  
•  MCF10A	
  	
  
      –  Normal	
  breast	
  cancer	
  
             •  Used	
  by	
  Horizon	
  Dx	
  to	
  produce	
  isogenic	
  cell	
  lines	
  with	
  cancer	
  relevant	
  muta*ons	
  
Other	
  
•  African	
  American	
  sample	
  with	
  70%	
  sanger	
  sequence	
  (ID?)	
  
      –  No	
  cell	
  line	
  available	
  
      –  Subject	
  s*ll	
  alive	
  =>	
  re-­‐consent	
  &	
  generate	
  cell	
  line?	
                                               | 16
Reference	
  Material	
  PorVolio	
  
Synthe*c	
  DNA	
  
•  What	
  types	
  are	
  most	
  useful	
  
     –  Variant	
  types,	
  sequence	
  context,	
  phasing	
  
           •  Assess	
  strengths	
  &	
  weaknesses	
  of	
  plaVorm	
  
           •  Test	
  valida*on	
  for	
  exis*ng	
  and	
  medically	
  relevant	
  variants	
  
•  What	
  size	
  
•  How	
  many	
  synthe*c	
  RMs	
  
     –  PlaVorm	
  valida*on	
  
     –  Test	
  valida*on	
  
•  Which	
  exis*ng	
  sources	
  should	
  be	
  considered	
  
     –  Available	
  for	
  research	
  and	
  for	
  profit	
  
•  Source	
  of	
  commercial	
  development	
  and	
  distribu*on	
  
     –  Manufactured	
  under	
  quality	
  system	
  for	
  diagnos*c	
  applica*ons	
  


                                                                                                    | 17
Reference	
  Material	
  PorVolio	
  

                                                                         n ce
Synthe*c	
  DNA	
  
•  What	
  types	
  are	
  most	
  useful	
  
      –  Variant	
  types,	
  sequence	
  context,	
  phasing	
   n
                                                        -co         fere e
                                   by        tele                 artic ipat
            •  Assess	
  strengths	
  &	
  weaknesses	
  of	
  plaVorm	
  

     What	
  size	
   vie
                               w                        to p
            •  Test	
  valida*on	
  for	
  exis*ng	
  and	
  medically	
  relevant	
  variants	
  
• 
            Resynthe*c	
  Rnvited iMs	
  
                       are
•    How	
  many	
  
                   ll
             Aalida*on	
  
      –  PlaVorm	
  valida*on	
  
      –  Test	
  v
•  Which	
  exis*ng	
  sources	
  should	
  be	
  considered	
  
      –  Available	
  for	
  research	
  and	
  for	
  profit	
  
•  Source	
  of	
  commercial	
  development	
  and	
  distribu*on	
  
      –  Manufactured	
  under	
  quality	
  system	
  for	
  diagnos*c	
  applica*ons	
  


                                                                                                     | 18
Confidential – Do not copy or distribute   | 19
Divider slide

Reference Materials Selection and Design Working Group Summary Aug2012

  • 1.
    Next  Genera*on  Sequencing:   Reference  Material  (RM)  Selec*on  and   Design   Workgroup   Genome  in  a  BoAle  Consor*um  /  NIST   August  16-­‐17,  2012  
  • 2.
    RM  Scope   • Human  Genome  &  Tumor  Sequencing   •  Variant  Types   –  SNP   –  InDel  /  Subs*tu*on   –  CNV   –  Structural  variant   | 2
  • 3.
    Reference  Material  Needed  For   •  PlaVorm  valida*on   –  Sequencing  System   –  Bioinforma*cs/Analysis  Pipeline   •  Test  valida*on   –  Whole  genome   –  Targeted   –  Germline  vs.  tumor   •  Reagent  valida*on   | 3
  • 4.
    New  York  State  proposed  dra[   guidelines  for  NGS  clinical  cancer  tests     •  …  A  posi%ve/sensi%vity  control  should  be  included  to   verify  analy*c  sensi*vity.  We  suggest  including  an   individually  bar  coded  low  posi*ve  control  near  the   sensi*vity  of  the  assay,  containing  mul*ple  known  soma*c   altera*ons  of  each  kind  to  be  detected,  to  verify  that  low   percentage  gene*c  variants  can  be  detected  in  every  run.  …   •  Performance  characteris*cs  for  each  type  of  gene*c   altera*on  the  assay  is  intended  to  detect,  e.g.  SNVs,  indels,   CNVs,  must  be  established  and  validated   •  Sequence  a  well-­‐characterized  reference  sample  (e.g.   HapMap  DNA  GM12878)  to  determine  error  rate  and   establish  depth/uniformity  of  coverage  across  all   amplicons.   | 4
  • 5.
    Associa*on  for  Molecular  Pathology   Comments  to  FDA  UHT-­‐Sequencing  Mee*ng,  June  2011   •  …  Performance  of  and  coverage  needs  for  a  given  plaVorm   are  likely  to  differ  depending  on  the  nucleic  acid  and  DNA   regions  analyzed,  the  variants  interrogated,  the  rela*ve   allele  propor*ons  of  par*cular  variants,  …  Evalua*on   should  consider  the  effects  of  rela*ve  GC  content,   homopolymeric  and  other  regions  of  repe**ve  sequence,   homologous  gene  regions  and  DNA  structural  variants,  …   This  necessitates  flexibility  and  individualiza*on  in  the   development  of  valida*on  protocols,  guidelines,  and   controls  on  a  (clinical)  applica*on-­‐by-­‐applica*on  basis.  …     •  Assay  controls  should  include  a  range  of  variants,  …   Process  controls  like  NA12876  …  and  the  synthe*c  ERCC   RNA  transcripts  from  NIST  are  examples  of  poten%al   standard  reference  materials.  …   | 5
  • 6.
    Reference  Material  PorVolio   Whole  Human  Genome   •  What  sources  of  RMs  to  consider   –  Available  for  research  and  for  profit   –  Primary/  cell  line   •  What  extent  of  prior  characteriza*on   •  Which  ethnici*es,  genders   •  Which  muta*ons  need  to  be  present   –  Is  medical  relevance  necessary   •  Ini*ally  to  have   –  ONE  characterized  genome  RM      -­‐  or   –  Mul*ple  genomes,  lower  level  of  characteriza*on   •  Source  of  commercial  development  and  distribu*on   –  Manufactured  under  quality  system  for  diagnos*c  applica*ons   | 6
  • 7.
    Workgroup  AAendees   • Approximately  25  aAendees   –  Federal,  incl.  FDA,  CDA,  NIST   –  Lab  accredita*on   –  Clinical  reference  labs   –  PlaVorm  technologies   –  Reference  material  /  reagent  providers   –  Research   | 7
  • 8.
    Workgroup  Summary  (1)   Reference  Material  Use   •  Characterize  PlaVorms  &  Methods     –  DNA  sequencing   –  Exis*ng  &  upcoming  NGS  technologies   –  Research  applica*ons   –  PlaVorm  &  methods  characteriza*on  for  clinical   diagnos*cs  applica*ons   •  Not  intended  as  reference  material  for   –  Valida*on  of  specific  muta*ons  in  a  panels   Confidential – Do not copy or distribute | 8
  • 9.
    Workgroup  Summary  (2)   Selected  Sample   •  NA12878     –  6,000  controlled  aliquots  of  10ug  each  on  order  by   NIST  from  Coriell   | 9
  • 10.
    Workgroup  Summary  (2)   Selected  Sample   •  NA12878     –  6,000  controlled  aliquots  of  10ug  each  on  order  by   NIST  from  Coriell   HOWEVER …. | 10
  • 11.
    Workgroup  Summary  (3)   Consent   •  Consent  available  for     –  Research    ü   •  Consent  not  available  for   –  Commercial   •  Incl.  altera*ons,  re-­‐distribu*on   –  Re-­‐iden*fica*on  through  sequence  data   NA12878 and family requires re-consenting for commercial use and potential re-identification through sequence data | 11
  • 12.
    Workgroup  Summary  (4)   Sample  &  Consent   •  If  re-­‐consen*ng  of  NA12878  not  possible,   replace  with  other  sample  (Trio)   •  If  necessary,  do  sample  change  now,  since   early  in  process  of  selec*ng  reference  sample   •  New  characteriza*on  of  reference  material   will  be  in  much  greater  depth  and  with   technologies  that  are  more  advanced  than   exis*ng  data  for  NA12878   | 12
  • 13.
    Workgroup  Summary  (5)   Sample  Characteris*cs   •  Sample  characteris*cs  are  more  important   than  selec*on  of  specific  sample  IDs   •  More  reference  samples  preferred  over  fewer   samples   –  Prefer  8  fully  characterized  samples  at  high  depth   and  corresponding  trios  at  lower  depth  over  4   fully  characterized  samples  plus  trios   | 13
  • 14.
    Workgroup  Summary  (5)   Sample  Characteris*cs  (cont.)   •  High  Priority   –  Mul*ple  ethnici*es   •  Diversity  in  structural  varia*on  to  stress  systems   –  Balanced  female  to  male  ra*o   –  Cell  line,  low  passages   •  Replenish  supply   •  Nice  to  have   –  Interracial  marriage  samples     •  Controlled  admixture   •  Less  cri*cal   –  Phenotypic  characteriza*on   •  Reference  material  not  for  discovery   –  Access  to  RNA  or  *ssues   •  No  limitless  supply  of  material  with  iden*cal  characteris*cs   | 14
  • 15.
    Workgroup  Summary  (6)   DNA  RM  Requirements   •  DNA  from  low  passage  cell  lines   –  New  batches  from  low  passage  aliquots   •  Modify  DNA  purifica*on  in  future  to  keep  step  with  new  NGS   technologies   –  Current  purified  DNA  fragment  sizes  are  80-­‐100kb   •  OK  for  exis*ng  technologies   –  New  nanopore  technologies  may  need  Mbp  fragments   •  Agarose  embedding  is  proven  extrac*on  technology   •  Consider  footprint  analysis  of  all  batches  prior  to  distribu*on   –  Iden*fy  gene*c  dri[,  mix  ups,  ….  ,  develop  benchmarks   •  Reference  material  that  mimics  tumor  sample  characteris*cs   –  FFPE  embedded  cells?   | 15
  • 16.
    Workgroup  Summary  (7)   Sample  Source  Sugges*ons   Most  support   •  Personal  Genome  Project  Samples     –  Includes  trios     –  Use  sequence  data  to  derive  admixture   –  hAp://www.personalgenomes.org   –  Consent  includes  research  &  commercial  use  and  re-­‐iden*fica*on     Some  support  (if  consent  sufficient)   •  HS1011   –  Charcot  Marie  Tooth  cell  line   •  Lupski  et  al,  NEJM  2010   •  MCF10A     –  Normal  breast  cancer   •  Used  by  Horizon  Dx  to  produce  isogenic  cell  lines  with  cancer  relevant  muta*ons   Other   •  African  American  sample  with  70%  sanger  sequence  (ID?)   –  No  cell  line  available   –  Subject  s*ll  alive  =>  re-­‐consent  &  generate  cell  line?   | 16
  • 17.
    Reference  Material  PorVolio   Synthe*c  DNA   •  What  types  are  most  useful   –  Variant  types,  sequence  context,  phasing   •  Assess  strengths  &  weaknesses  of  plaVorm   •  Test  valida*on  for  exis*ng  and  medically  relevant  variants   •  What  size   •  How  many  synthe*c  RMs   –  PlaVorm  valida*on   –  Test  valida*on   •  Which  exis*ng  sources  should  be  considered   –  Available  for  research  and  for  profit   •  Source  of  commercial  development  and  distribu*on   –  Manufactured  under  quality  system  for  diagnos*c  applica*ons   | 17
  • 18.
    Reference  Material  PorVolio   n ce Synthe*c  DNA   •  What  types  are  most  useful   –  Variant  types,  sequence  context,  phasing   n -co fere e by tele artic ipat •  Assess  strengths  &  weaknesses  of  plaVorm   What  size   vie w to p •  Test  valida*on  for  exis*ng  and  medically  relevant  variants   •  Resynthe*c  Rnvited iMs   are •  How  many   ll Aalida*on   –  PlaVorm  valida*on   –  Test  v •  Which  exis*ng  sources  should  be  considered   –  Available  for  research  and  for  profit   •  Source  of  commercial  development  and  distribu*on   –  Manufactured  under  quality  system  for  diagnos*c  applica*ons   | 18
  • 19.
    Confidential – Donot copy or distribute | 19
  • 20.