Routine Use of ForenSeq Solution on Casework Samples: Feedback One Year After Lab Implementation

Transcription

1 Routine Use of ForenSeq Solution on Casework Samples: Feedback One Year After ab Implementation ille DEII Barcelona Paris yon François-Xavier AURENT, Ph.D Head of R&D in Forensic Genetics Institut National de Police Scientifique Toulouse Marseille

2 Actual DNA workflow at INPS Reference samples (Database) Automated Genotyping Unit (yon only) FTA cards / year (double analysis) Casework samples Volume crime (swabs) Criminal cases + Mitochondrial DNA DNA phenotyping Synthetic DNA (spray/ink) DNA samples / year

3 Actual DNA workflow at INPS A DNA Extraction A DNA Quantification PCR mini-str PCR Y-STR PCR Autosomal STR PCR mitochondrial PCR Phenotypic SNP A : Automation : IMS (aboratory Integrated Management System) A A A A A CE CE CE Sanger CE Analysis Analysis Analysis Analysis Analysis R E P O R T Is it worth adding another workflow?

4 What we expected back in 2015! Our MPS workflow philosophy Premium workflow to allow a deeper analysis of complex DNA samples (degraded, mixtures ) Not to fully replace classic workflow (PCR-CE) Reserved only for important casework (serious crimes) Replace existing workflow (Yfiler, minifiler)? Automation and IMS ISO Major goals : Help experts to interpret genetic profiles Bring more caseworks to the lab and become expert in complex analyses Re-open cold cases

5 European survey on forensic applications of massively parallel sequencing (2017) Additionnal questions with regards to massively parallel sequencing implementation Automation? Equipment? Training? Genetic profile validation and interpretation? Developmental validation? Sample Tracking? Conclusions : Money, complexity, lack of feedback A. Alonso, P. Muller,. Roewer, S. Willuweit, B. Budowle, W. Parson. European survey on forensic applications of massively parallel sequencing. Forensic Sci. Int. Genet., 29 (2017), pp. e23-e25 Accreditation? Casework application?

6 Fully automated DNA library preparation TM Automation of Verogen ForenSeq ibrary preparation kit from PCR amplification to library pooling 8 to 96 samples Hamilton ID NGS-V STARlet Hands-on time from reduced from 3 hours to 15 minutes DNA analysis to data interpretation in 4 days Core Gripper Magnet + Heater Shaker F.X. aurent,. Ausset, M. Clot, S. Jullien, Y. Chantrel, C. Hollard,. Pene, Automation of library preparation using Illumina ForenSeq kit for routine sequencing of casework samples, Forensic Sci. Int. Genet. Suppl. Ser. 6 (2017) e415 e417

7 Developmental validation of ForenSeq solution with automated protocol Validation guidelines issued by the Scientific Working Group on DNA Analysis Methods (SWGDAM) in 2012 available here : Revised in November 2016 to address Next Generation Sequencing (NGS) technologies Parameters tested in our laboratory : Sensitivity (full profile from 250 pg as DNA input) Repeatability (CV < 10%) Reproducibility (CV < 20%) Concordance (99,54 % compared to GlobalFiler) Cross-contamination Publication under review at FSI Genetics Mixture studies (220 samples) ongoing

8 Decision tree for marker validation rules

9 Improvements to maximize MPS usage in a forensic laboratory Pooling library adjustment Combined analysis of rich and poor DNA samples in the same run DNA sample concentration > ng/µ

10 Improvements to maximize MPS usage in a forensic laboratory Pooling library adjustment Pooled library quantification Qubit 3.0 (ThermoFisher Scientific)

11 Improvements to maximize MPS usage in a forensic laboratory Pooling library adjustment Pooled library quantification Sample tracking File exchange concept

12 - Global feeling is hard to apprehend with the actual visualisation tool discard genetic profiles that do not meet basic criteria at a glance focus on complex profiles that requires careful attention - Needs of additional thresholds (ex: threshold for homozygous loci) - User modifications tracking (sample history available on UAS v1.3) - Numerous mouse-clicks UAS still needs some improvement for a routine use Feedback from the forensic community will be determinant to improve analysis software and adapt them to the needs of forensic laboratories

13 Validation rules automatically applied - Green : validated - Orange : ACR issue - Red : non valid Institut national de police scientifique Development of PIRANHA Programme d Interprétation Résultats d Analyses NGS Hautement Amélioré R shiny web application Data from the UAS summary report (.xlsx file) Number of reads reported for each allele

14 autosomal STR results Institut national de police scientifique

15 X and Y STR results Institut national de police scientifique

16 Sequences alignment for viewing isomutations

17 Improvements to maximize MPS usage in a forensic laboratory Pooling library adjustment Pooled library quantification Sample tracking IMS integration Inspired from PA. Bario, P. Martin, The DNASEQEX Consortium, A. Alonso. IMS configuration to fit new massively parallel sequencing workflows in forensic genetics. Forensic Sci. Int. Genet. Suppl. Series, 29 (2017), pp. e104-e106

18 Feedback on caseworks samples More than 300 DNA extracts analyzed (in parallel to GlobalFiler analysis) DNA concentration: 0, ng/µ Why the DNA extracts were analyzed with ForenSeq? - To complete a genetic profile obtained from a highly degraded DNA - To identify a very minority contributor - To deconvoluate mixture with the help of possible isomutations - To have additional markers for complex kinship

19 Full profiles with GlobalFiler hb <0,5 nb of reads < homo it at least 1 allele <it discordance

20 Troubles with D22S1045: «The autosomal STR locus D22S1045 should be interpreted with caution as more imbalance in read counts may be seen between alleles of a heterozygote than observed at other loci» Jäger et al. FSI Genetics

21 Partial profiles with GlobalFiler discordance + GF + FSQ

22 «Dominant» profiles with GlobalFiler discordance + GF + FSQ

23 discordance Mixture + GF + FSQ Isomutation At least 2 isomutations per mixture

24 SNP usage for kinship analysis on postmortem identification 46 SNP with one common allele 0 discordance ikelihood Ratio (R) with Familias : - 20 STR GlobalFiler (1 discordance) : STR GlobalFiler (1 discordance) + 4 STR ForenSeq : STR GlobalFiler (1 discordance) + 4 STR ForenSeq + 46 SNP ForenSeq : Allelic frequencies obtained from 1000 genomes project Still questioning if it's statistically ok? Independance of SNPs and STRs?

25 First MPS DNA profile sent to the French National DNA Database in February 2018 Cold Case of 2011 (Homicide) First analyzes in the laboratory of Paris (Identifiler) 2 DNA extracts (mixture of the victim + a very minor male contributor) Minor Contributor? Mixture Minor? Mixture Victim CE? NGS Confirmation of the CE results thanks to the two «17» Validation of 4 additional markers and 5 alleles per unknown DNA profile One new genetic profile registered to the national DNA database One partial genetic profile completed 2 isoalleles

26 Future DNA analysis workflow in yon

27 Future DNA analysis workflow in yon From January 2019 : - Direct use of MPS Body identification through kinship analysis (outside of parent-child link) - If necessary after STR-CE : Vaginal swab sample from rape case (without male profile) Degraded DNA (unique profile) Mixtures (2-3 contributors) Half profile (foetuses from rape) Rare alleles caracterization

28 Acknowledgment aurence Devesse Yann egros Nicola Oldroyd Clark Patrick Sageat Kameran Wong ionel Ausset Gianluca Carboni Yann Chantrel Yann Chovory Maeva Clot Stéphanie Fauroux Clémence Hollard Sandrine Jullien Agnès Milon aurent Pène Élodie Suzanne