Challenges and Paradigm Shifts by the Adoption of MPS in Forensic Casework. Lessons Learned from the Collaborative DNASeqEx Project so far.

Transcription

1 Challenges and Paradigm Shifts by the Adoption of MPS in Forensic Casework. Lessons Learned from the Collaborative DNASeqEx Project so far. Sascha Willuweit, Charité - Universitätsmedizin Berlin, Institute of Legal Medicine, Forensic Genetics Dept. on behalf of the DNASeqEx Project

2 Summarize MPS Challenges Lack of Nomenclature National DBs incompatibility Missing Population Data Lack of Legislation Missing Validation Guidelines No Proficiency Test Not Reliable 0 Data from A. Alonso, et al., European survey on forensic applications of massively parallel sequencing, Forensic Sci. Int. Genet. (2017)

3 Summarize MPS Challenges Lack of Nomenclature National DBs incompability Missing Population Data Lack of Legislation Missing Validation Guidelines No Proficiency Test Not Reliable itself (and the understa 0 Data from A. Alonso, et al., European survey on forensic applications of massively parallel sequencing, Forensic Sci. Int. Genet. (2017)

4 Challenges / Topics of this Talk Technology / Level of Nomenclature

5

6 1µl of 1ng/µl DNA [650 g/mole molecular weight] [6.48 pg per genome] 600 single-stranded copies

7 targets (STRs, SNPs, )

8 600 single-stranded copies [approx. 50 markers] 25 PCR cycles 1 trillion target copies [approx. 200bp length per marker/copy] 200 trillion target base-pairs 100:1 target DNA:genome DNA

9 1 trillion target copies [50 samples(1ng) pool in library] 50 trillion target copies 25 million good cluster [x2 selecting orientation] [x2 heterogeneous clusters] itial 100 million sequences bound 50,000:1 ratio available/bound

10 initial 100 million sequences 1:2 chance of 25th-cycle-copy 3 billion chance of 1st-cycle-cop

11 25 million reads [50 samples] ~500,000 reads per sample [50 markers] 0,000 reads per sample/marke [150 markers] 3000 reads per sample/marker

12 Errors PCR (Taq) ~ 10 12% PCR (Pfu, Pwo) ~ 1 2% Cloning ~1%

13 nucleotide transition predominate huge impact on MPS : we are sequencing PCR (Taq) ~ 10 12% PCR (Pfu, Pwo) ~ 1 2% Cloning ~1%

14 1000 copies per cluster/read may invalidate a cluster PCR (Taq) ~ 10 12% PCR (Pfu, Pwo) ~ 1 2% Cloning ~1%

15 3% resulting error rate (theoreti orithms allowing for k-mismatch don t hide artifacts

16 Artifact Example A: 160 reads B: 16 reads C: 4 reads D: 3 reads E: 2 reads F: 1 read G: 1 read threshold 3 reads

17 Artifact Example A: 160 reads B: 16 reads C: 4 reads D: 3 reads

18 Artifact Example A: 160 reads B: 16 reads C: 4 reads D: 3 reads A is most likely to be my sequence

19 Artifact Example A: 160 reads B: 16 reads C: 4 reads D: 3 reads E: 2 reads F: 1 read G: 1 read no threshold

20 Artifact Example A: 160 reads B: 16 reads C: 4 reads D: 3 reads E: 2 reads F: 1 read G: 1 read

21 Artifact Example A: 160 reads D: 3 reads C: 4 reads B: 16 reads E: 2 reads F: 1 read G: 1 read

22 Artifact Example A: 160 reads D: 3 reads C: 4 reads B: 16 reads E: 2 reads F: 1 read G: 1 read A and B are (possibly) valid sequences ANALYSIS SOFTWARE SHOULD NOT JUDGE HERE

23 Absolute Reads Example lab A: avg. reads 309.1±264 lab B: avg. reads 787.4±536.3

24 Absolute Reads Example lab A: avg. RRU 28.1±24 lab B: avg. RRU 25.4±17.3

25 Nomenclature Assumptions (A)full sequence data is neither feasible nor necessa (B)sequence variants are countable (not infinite) (C)there are stable variants (D)exact sequence information is available (E) there is no molecular correct and easy humanreadable syntax to name variants

26 Nomenclature Assumptions (A)full sequence data is neither feasible nor necessa full sequence > information not human-graspable or -comparable 3 suspects, 10 mixed stains, 50 markers approx. 300 pairs of 200±100 bases

27 Nomenclature Assumptions (B)sequence variants are countable (not infinite) finite sequence mutation rate << 1 error rate << 1

28 Nomenclature Assumptions (C)there are stable variants need to recognize variants allele space has major variants (population stud

29 Nomenclature Assumptions (D)exact sequence information is available access to sequence is given primer sequences can be removed (thus are know

30 Nomenclature Assumptions (E) there is no molecular correct and easy humanreadable syntax to name variants primer independent position(s) of delta(s) reference genome name/version covered region

31 Nomenclature Proposal: NOMAUT (under development) basic principle of a catalogue/oracle takes: marker, sequence, region, orientation returns: CE-allele + catalogue designation self-maintained system, because querying the database will create temporary variants (immanent) temporary variants become fixed variants by independent observations (new query, new lab) oracle: will handle comparisons as well not a software, but a service will be announced at the 27th ISFG congress

32 Summary / Software

33 Summary / Software relative read units are good for comparability

34 Summary / Software easy nomenclature to handle MPS data

35 Summary / Software don t hide (minor) sequences / artifacts

36 Summary / Software don t hide sequence information

37 Summary / Software PLEASE! Don t over-simplify MPS to CE-mimic bar plots.

38 Thanks DNASeqEx project group Antonio Alonso Alonso Pablo Martin Martin Pedro Barrio Caballero Walther Parson Burkhard Berger Petra Müller Lutz Roewer Sascha Willuweit Steffi Köcher Bruce Budowle

39 Speaker was provided travel and hotel support by Thermo Fisher Scientific for this presentation, but no remuneration When used for purposes other than Human Identification or Paternity Testing the instruments and software modules cited are for Research Use Only. Not for use in diagnostic procedures. Thermo Fisher Scientific and its affiliates are not endorsing, recommending, or promoting any use or application of Thermo Fisher Scientific products presented by third parties during this seminar. and materials presented or provided by third parties are provided as-is and without warranty of any kind, including regarding intellectual property rights and reported results. Parties presenting images, text and material represent they have the rights to do so.