Investigating the use of MPS for non-human and other specialist forensic applications

Transcription

1 Investigating the use of MPS for non-human and other specialist forensic applications Jon Wetton DNA Evidence to Investigative Insights: Illumina - Berlin 19 th May 2016

2 MPS approaches to both human and animal forensic genetics Increased discrimination power More distinct variants Better for degraded DNA No need to ensure non-overlapping size ranges of STRs Mini-STRs & short amplicon SNPs Simultaneously target STRs & SNPs both nuclear & mtdna Saves time, money and scarce sample Opportunity to exploit new forensic markers & applications Examples from cats, dogs and humans Evaluating new technologies Creating reference databases Developing novel profiling & interpretation methods Testing in realistic scenarios Validation for casework implementation

3 Novel applications and the need for databases: A casework example S u s p e c t V i c t i m

4 The pink shower curtain Curtain examined by LGC Tinker (suspect s cat) LouLou (victim s cat)

5 Comparing national mitochondrial DNA databases US lab profiled mtdna HV 493 US cats found plus Europe, Middle East, Asia 25 populations, 22 breeds 12 common types Tinker Number of mutational differences cat 2 cats 10 cats Figures modified from Grahn et al

6 Building a representative cat mtdna database

7 Comparing national mitochondrial DNA databases US lab profiled mtdna HV 493 US cats found plus Europe, Middle East, Asia 25 populations, 22 breeds 12 common types UoL - profiled 152 cats 32 local, 120 rest of England C-UK2 C-UK1 B-UK1 Tinker B-UK2 Number of mutational differences cat 2 cats 10 cats U

8 Comparing national mitochondrial DNA databases

9 Comparing the local and national datasets

10 Determining the evidential strength Dependant upon rarity of profile and size of database Table of Verbal Equivalents Limited evidence to support LR <1-10 Moderate evidence to support LR Moderately strong evidence to support LR Strong evidence to support LR Very strong evidence to support LR >10000 Supported by other transfer evidence from the shower curtain to the suspect s clothes

11 Improving the discrimination of the mtdna database Police fortunate due to rarity of Tinker B-UK1 DNA type 10 million cats in UK are tagging items in a quarter of households US test examined less than 3% of mtdna, MPS gives access to the rest UK Database

12 UK database (n=152) UK Sanger HVSI database Cat MPS database

13 MPS UK UK Sanger HVSI database database HVSI (n=152) database (N=35) Cat MPS database

14 Cat MPS database MPS UK Sanger UK HVSI database HVSI database database (n=152) (N=35) Whole mtdna database A (4 out of 5 distinguishable), B (8/8), B1 (2/2), C (3/3), E (2/2) but D, F & I no additional variation D at high frequency in South Asian breed cats, two of three tested are Burmese & Tonkinese Random Match Probability reduced from 0.19 to 0.05 Control Region only - just 5 out of 35 were unique but with whole mtdna 23 out of 35 unique Testing additional B-UK1 revealed a further SNP splitting them into two variants

15 Building a representative dog mtdna database Dog sample origins Commonest dog breeds Labrador Jack Russell Terrier Staffordshire Bull Terrier Border Collie Greyhound

16 B1 25% number B2 1% B3 x x x 1% Dog x MPS x mtdna database 0% 0% 0% 0% 0% x C2 A1 0% x 3% 0% 9% A2 6% A3 1% A4 6% x 0% x 0% x A5 0% 1% A6 1% A13 3% x 0% x 0% x 0% x 0% A10 1% A12 13% A9 1% A11 3% x 0% A8 1% x 0% A7 22% Dogs of the same breed are more likely to match than random dogs but there is variation Random Match Probability reduced from 0.15 to 0.03 Control Region only - just 8 out of 69 were unique but with whole mtdna 38 out of 69 unique

17 Amplicon Lenth (bp) Whole mitochondrial sequencing from degraded hair DNA Shed hairs contains virtually no nuclear DNA, but some mtdna DNA is both limited and degraded 63 overlapping regions targeted DNA replicated in 6 plexes, each with input one 14 yr old guard hair Complete coverage of mtdna Lower limit of sensitivity yet to be established

18 ForensiCats: How cats can solve crimes Experiment.com

19 Human Y chromosome markers Y chromosome SNPs PAR1 NRY PAR2 SNPs C T A T G C T A G C T G T A C G G Haplogroup STRs Haplotype

20 Association between haplogroup and haplotype MPS data derived from 448 globally sampled males revealed 13,000 new Y SNPs, potential forensic use Hallast, P. et al. The Y-chromosome tree bursts into leaf: 13,000 high confidence SNPs covering the majority of known clades. Mol. Biol. Evol. 32, (2015). Batini C et al Large-scale recent expansion of European patrilineages shown by population resequencing. Nature Communications DOI: /ncomms8152 (2015) DYS19 DYS389I DYS389II DYS390 DYS391 DYS392 DYS393 DYS385a DYS385b DYS437 DYS438 DYS439 DYS448 DYS456 DYS458 DYS635 YGATAH4 DYS481 DYS533 DYS549 DYS570 DYS576 DYS643 R1b R1a I1 J1

21 Association between haplogroup and haplotype Number of mutational differences man 2 men 10 men Abergele (26) Llangefni (32) Llanuwchllyn (30) Tregaron (30) Chippenham (55) Sussex (26) Winchester(36) Friesland (91)

22 Haplogroup predictions of UoL dataset E I2 R1b I1 G R1a J2

23 Variation among Friesland samples Abergele (26) Llangefni (32) Llanuwchllyn (30) Tregaron (30) Chippenham (55) Sussex (26) Winchester(36) Friesland (91)

24 Variation among English samples Abergele (26) Llangefni (32) Llanuwchllyn (30) Tregaron (30) Chippenham (55) Sussex (26) Winchester(36) Friesland (91)

25 Variation among Welsh samples Abergele (26) Llangefni (32) Llanuwchllyn (30) Tregaron (30) Chippenham (55) Sussex (26) Winchester(36) Friesland (91)

26 Variation among Welsh samples (- Abergele) Abergele (26) Llangefni (32) Chippenham (55) Sussex (26) Llanuwchllyn (30) Winchester(36) Tregaron (30) Friesland (91) Hence forensic databases must be geographically representative for near matches

27 Geographic variation among Saudi males Central SA (~8 million, N=40) South SA (~4.2million, N=54) J1 Data from Yahya Khubrani, Saudi PhD student

28 Geographic variation between socially isolated populations - Sinai Bedouin Mzeina (~16,000, N=29) Jebeliya (~5,000, N=25) J1 One PPY23 profile shared by 10 Mzeina men (34% of sample) Three PPY23 profiles each shared by 5 Jebeliya men (20% each) Hence forensic databases must be representative of relevant communities even when living together Data from Katalina Bobowik, visiting MSc. student

29 High frequency haplotypes in Ireland A Y-Chromosome Signature of Hegemony in Gaelic Ireland Moore et al. American Journal of Human Genetics Volume 78 February 2006

30 Coinheritance of Y-STR profiles and surnames A B C D E F G Marker Allele Allele Allele Allele Allele Allele Allele 1 DYS DYS389I DYS389II DYS DYS DYS * * * * 7 DYS DYS DYS DYS DYS DYS DYS DYS DYS DYS GATA H Sources of error multiple founders (Smith ), illegitimacy, adoption False positive rate significantly reduced by more discriminating Y-STR test STR sequence variants might provide more robust links than length Wetton Wetton variants

31 People of the British Isles The Impact of Diasporas on the Making of Britain Aim Investigate sex-biased migration within the British Isles mtdna (n=2836) Illumina and Affy Y-DNA (n=1073) Illumina and Affy6 Data Grouped into thirty sampling areas (Counties) Leslie S. et al. (2015) The fine-scale genetic structure of the British population Nature 519: (19 March 2015) DOI: /nature14230

32 Distribution of specific Y chromosome variants in the British Isles

33 People of the British Isles BBSRC icase studentship collaboration MPS Sequencer (Illumina MiSeq) 1) MPS Kit (Promega PowerSeq) Concordance Y-STRs Autosomal STRs mtdna (PPY23) (PP Fusion) (HVSI & II) Databasing Mutation mechanisms & rates Geographic distribution 2) MPS Kit (Illumina ForenSeq) Concordance Phenotypic SNPs (Face of Britain) STRs (PowerSeq) Databasing In addition Y-SNPs, Autosomal SNPs, X-STRs

34 Increased Discrimination from MPS TAGA/TACA DYS635 DYS393 ATCT/ATCG Split iso-alleles Significant sequence changes are phylogenetically informative Forensic Y-STR locus choice excluded many complex loci More informative MPS loci could be incorporated Knowledge of allele length / haplogroup correlation may help with deconvolution of mixtures A1b B2a D2 E-V38 E-M78 E-M81 I1 I1 I1 I1 I2 J1 J2 J2 L N O2a R1a R1a R1b R1b R1b R1b R1b Data from Tunde Huszar, UoL/KFS icase student

35 Summary MPS offers new investigative avenues for haploid markers Population relevant databases are key to ensure the evidential weight of a rare variant is appropriately assessed Evidence of identity / very close relationship or Localized frequency hotspot

36 Gurdeep Matharu Lall, Barbara Ottolini, Tunde Huszar, Rita Neumann, Chiara Batini, Emma Sterling, Emma Walklett, Reshma Vaghela, Nic Sylvius, Rita Rasteiro, Celia May, Marwan El Khoury & Mark Jobling Human DNA PhD projects Acknowledgments Animal DNA Deb Hopwood (LGC)