Vojtěch Bystrý CEITEC Bioinformatics Core Facility CEITEC research areas FI MUNI Bioinformatics Seminar 2 Bioinformatics CF mission Primary NGS analysis S econdary standard NGS analysis Custom project specific analysis ID: 1045592
Download Presentation The PPT/PDF document "NGS applications in molecular medicine" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. NGS applications in molecular medicineVojtěch Bystrý
2. CEITEC Bioinformatics Core FacilityCEITEC research areasFI MUNI Bioinformatics Seminar2
3. Bioinformatics CF mission Primary NGS analysisSecondary standard NGS analysisCustom (project specific) analysisBioinformatics educationGCFFI MUNI Bioinformatics Seminar3
4. Bioinformatics CF mission Primary NGS analysisSecondary standard NGS analysisCustom (project specific) analysisBioinformatics educationGCFAUTOMATEFI MUNI Bioinformatics Seminar4
5. Bioinformatics CF mission Primary NGS analysisSecondary standard NGS analysisCustom (project specific) analysisBioinformatics educationGCFFI MUNI Bioinformatics Seminar5
6. Primary NGS analysisInformation system for NGS sequencing and standard analysisNew system in development supported by TACR - TJ04000141IT infrastructure maintenanceNew storage for sensitive dataTransfer of computations to HPCFI MUNI Bioinformatics Seminar6
7. Information system for NGS analysisLab technicianClinical geneticist/doctorBioinformaticianFI MUNI Bioinformatics Seminar7
8. Information system for NGS analysisLab technicianClinical geneticist/doctorBioinformaticianFI MUNI Bioinformatics Seminar8
9. Information system for NGS analysisFI MUNI Bioinformatics Seminar9
10. FI MUNI Bioinformatics SeminarInformation system for NGS analysisGoal is clinical specific results visualization10
11. Primary NGS analysis - infrastructurehdhu-clusterkubernetes clusterNeccesary with increased sequencing throughput FI MUNI Bioinformatics Seminar11
12. Primary NGS analysis - infrastructuredatastorageTask RunnerFI MUNI Bioinformatics Seminar12
13. Bioinformatics CF mission Primary NGS analysisSecondary standard NGS analysisCustom (project specific) analysisBioinformatics educationGCFFI MUNI Bioinformatics Seminar13
14. Secondary standard NGS analysisComprehensive repository of bioinformatics pipelineConstant workflow improvementHPC deployableFI MUNI Bioinformatics Seminar14
15. Secondary standard NGS analysisAnalysis2018201920202021RNA-Seq (mapping, QC, gene count)DNA re-sequencing (mapping, QC)SmallRNA-Seq (mapping, QC, gene count)Differential expression analysisAlternative splicing analysisGene ontology/ Pathway/ GSEA analysisVariant callingFusion genes detectionChip/CLIP-SeqTranscriptome assemblyCRISPR analysisCopy number analysisscRNA-primary analysismetagenomic 16S rRNA-SeqRNA-editing analysisnot availablein developmenttested / used in projects automated / standardisedFI MUNI Bioinformatics Seminar15
16. Bioinformatics CF mission Primary NGS analysisSecondary standard NGS analysisCustom (project specific) analysisBioinformatics educationGCFFI MUNI Bioinformatics Seminar16
17. NGS applications in molecular medicineComplex CFB project examples:A-C-G-T project - 1000 Czech genomes for better variant interpretation High-risk paediatric patients with solid tumorsRadiomics and Genomics combination for solid tumor diagnosticsFI MUNI Bioinformatics Seminar17
18. A-C-G-T: Analysis of Czech Genomes for Teranostic1000 Czechs from all regionsBlood sampleDNA isolationWGS sequencingACGTCGTACTAGCCCCCCAATTGGCCADB of germline allelic frequencies in czech population123456Genetic diagnostic improvment TERApy + diagNOSTICFI MUNI Bioinformatics Seminar18
19. A-C-G-T: BioinformaticsStandard germline variant calling analysisFI MUNI Bioinformatics Seminar19
20. A-C-G-T: Bioinformatics - Workflow selectionStandard germline variant calling analysisTested standard workflows:bcbioSnakemakeNextflownf-core/sarek https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7111497/FI MUNI Bioinformatics Seminar20
21. A-C-G-T: Bioinformatics - challengesScale of the projectSample and data organization and managementComputational resources300 TB of raw dataOne sample ~ 24 hour analysis on 16 CPU machineFI MUNI Bioinformatics Seminar21
22. A-C-G-T: Bioinformatics - challengesScale of the projectSample and data organization and managementComputational resources300 TB of raw dataOne sample ~ 24 hour analysis on 16 CPU machine~ 3 years analysis time on 16 CPU machine for 1000 samplesFI MUNI Bioinformatics Seminar22
23. A-C-G-T: Bioinformatics - infrastructure solutionelixir datastorageTask RunnerFI MUNI Bioinformatics Seminar23
24. Next stepselixir datastorageTask RunnerIt is not yet working as shown FI MUNI Bioinformatics Seminar24
25. Next steps – GPU accelerationelixir datastorageTask RunnerFI MUNI Bioinformatics Seminar25
26. High-risk paediatric patients with solid tumorslong term collaborative project prof. Slaby research group at CEITEC MU Department of Pediatric Oncology of The University Hospital Brno – prof. Sterbapatients with expected 5y overall survival less than 30%both newly diagnosed and relapsed ~ 50 patients are admitted to the program each yearFI MUNI Bioinformatics Seminar26
27. Personalized therapy in pediatric oncologyTumor samplescooperationgermline and somatic mutation analysis - WESfusion genes detection - targeted RNA-seqCGH arrays – CNVgene expression profiling – RNA-seqglobal methylation profiles - bisulfite sequencingFI MUNI Bioinformatics Seminar27
28. Case report5 years old boy with diffuse intrinsic pontine glioma (DIPG), 6 months of standard chemo/radiotherapy -> tumor progression, only 6 months to liveGenomic analysis identified activation mutation in PI3K kinase (Akt oncogenic signalling pathway)Miltefosin/impavido(only approved Akt inhibitor)DRUG REPURPOSINGLeishmaniasisAt the begginingAfter 6 month treatment4 month of Miltefosin8 month of MiltefosinFI MUNI Bioinformatics Seminar28
29. Personalized therapy in pediatric oncologyTumor samplescooperationgermline and somatic mutation analysis - WESfusion genes detection - targeted RNA-seqCGH arrays – CNVgene expression profiling – RNA-seqglobal methylation profiles - bisulfite sequencingMulti-omicsFI MUNI Bioinformatics Seminar29
30. Current bioinformatics analysisWES Somatic and germline variant callingMaximal sensitivityMaximal annotationCustom cancer DBsTMB computation – problematicFusion genesBest tool so far – ARRIBANeed visual confirmation IGVImproved workflows every yearFI MUNI Bioinformatics Seminar30
31. Overall statisticsWES2341217Fusionswith transcriptomics193FI MUNI Bioinformatics Seminar31
32. WES statisticsDiagnosesNumber of analysisActionable variantRelative actionable %Affected genesCNS tumor852428PIK3CA, PIK3R1, BRAF, FGFR1, KRAS, PDGFRA, EGFR, ACVRL1, SMARCB1Sarcoma76811NRAS, HRAS, NF1, JAK2Carcinoma1218NRASNeuroblastoma16531ALK, BRAF, FGFR1, NF1Hematooncological malignancies9111KRAS, JAK3Vascular anomalies8788TEK, PIK3CAHistocytosis44100BRAF, MAP2K1Fibromatosis11873CTNNB1, PDGFRB, EGFROther19421NF1, NF2, CTNNB1Total2406226FI MUNI Bioinformatics Seminar32
33. Fusion genes statisticsFusion groupNumber of findingsTypical diagnoseEWSR fusions19Ewing sarcomesBRAF fusions11Astrocitomas - glioma typeNTRK fusions6Infantile fibrosarcomePAX3-FOXO110ARMS SS18-SSX1/25RELA/YAP11 fusions (ependymoma-related)7BEND2-MN12TGF-GPR1283other 8Total258positive for known fusions:71diagnostic significance:64actionable:22FI MUNI Bioinformatics Seminar33
34. Methylation analysisClassification of central nervous system tumorsClassifier from EMBL groupDesigned for chipsWe make it work for NGSNew WHO classification is in preparationHelp with diagnose if histopathology is problematic faster and cheaper than fusion, WESTry to develop our own algorithmsProblem is with training dataFI MUNI Bioinformatics Seminar34
35. Transcriptomic analysisRNA-seqOnly tumor sampleDifferential expression to reference DBGtexNot optimalSignificant overexpressionsOnly for genes in known onco-pathwaysSupporting evidenceLowest level of evidenceSupport other methods findings – WES, CNVBetter understatnding of the tumor molecular biologyImprove therapeutic regimesMetronomic therapyFI MUNI Bioinformatics Seminar35
36. Next steps Individual NGS analysisSpeed-up the processing for individual NGS analysisCGH arrays → low coverage WGSGeneralize and improve analysis for transcriptomics New models for methylations Incorporate individual NGS analysis (omics) results for one patientFI MUNI Bioinformatics Seminar36
37. Next steps - incorporate per patient results Interactive dashboardPredictive modelFI MUNI Bioinformatics Seminar37
38. Radiomics and Genomics combinationCollaboration with labs from AKH WienChristian Doppler Laboratory for Applied MetabolomicsDivision of Nuclear Medicine, Department of Biomedical Imaging Goal is to combine nuclear medicine imaging with NGS data for better cancer diagnostics/treatment strategy decisionsMulti-omics approachIndividual projectsHead and neck squamous cell carcinoma – HNSCC – 2020/21Prostate cancer – PCa – 2021/22Lung Cancer – LCa – 2022/23 38FI MUNI Bioinformatics Seminar38
39. General project workflow39Primary tumor tissue from 62 HNSCC patientsFDG-PET/CT imagingDNA - whole exome sequencingSomatic variant analysisFI MUNI Bioinformatics Seminar39
40. Somatic variant NGS data analysisPrimary analysis and QCVariant calling Variant annotationVariant interpretation Clinical applicationFI MUNI Bioinformatics Seminar40
41. Somatic variant NGS data analysisPrimary analysis and QCVariant calling Variant annotationVariant interpretation Predictive modeling (Aggregated feature extraction)…Clinical applicationFI MUNI Bioinformatics Seminar41
42. Genomic variant predictive modeling Genomic variant data are very problematic for modelingEnormous feature space~ 100 000 featuresLimited number of data pointsOnly one predictive label per patientFeature selection/extractionIncrease number of samplesFI MUNI Bioinformatics Seminar42
43. Genomic variant predictive modeling Genomic variant data are very problematic for modelingEnormous feature space~ 100 000 featuresLimited number of data pointsOnly one predictive label per patientFeature selection/extractionIncrease number of samplesCurse of dimensionalityFI MUNI Bioinformatics Seminar43
44. Genomic variant predictive modeling Genomic variant data are very problematic for modelingEnormous feature space~ 100 000 featuresLimited number of data pointsOnly one predictive label per patientFeature selection/extractionIncrease number of samplesBiologically meaningful data extractionUsage of publicly available dataCurse of dimensionalityVariantsGenesPathwayFI MUNI Bioinformatics Seminar44
45. Genomic variant predictive modeling Pathway level “disruption” score from gene- and mutation-level scoresKEGG pathwaysMutation effect combination of CADD, EVE, Polyphen2 scoresFI MUNI Bioinformatics Seminar45
46. HNSCC project resultsFI MUNI Bioinformatics Seminar46
47. Next steps - Spatial transcriptomicsfrom: DOI:10.1038/s41467-018-04724-5FI MUNI Bioinformatics Seminar47
48. We have many more projectsWeighted gene co-expression analysisProteomics and transcriptomics results combinationTranscription factor enrichment analysismicro-RNA expression effect on tumor progresionAmino-acid conservation and phylogeny analysisCRISPR based projectsRNA-seq profiles correlationPolyuridylation quantification…48FI MUNI Bioinformatics Seminar48
49. We need help Primary NGS analysisSecondary standard NGS analysisCustom (project specific) analysisdatastorageTask RunnerAnalysis2018201920202021RNA-Seq (mapping, QC, gene count)DNA re-sequencing (mapping, QC)SmallRNA-Seq (mapping, QC, gene count)Differential expression analysisAlternative splicing analysisGene ontology/ Pathway/ GSEA analysisVariant callingFusion genes detectionChip/CLIP-SeqTranscriptome assemblyCRISPR analysisCopy number analysisscRNA-primary analysismetagenomic 16S rRNA-SeqRNA-editing analysisFI MUNI Bioinformatics Seminar49
50. We need help Primary NGS analysisSecondary standard NGS analysisCustom (project specific) analysisdatastorageTask RunnerAnalysis2018201920202021RNA-Seq (mapping, QC, gene count)DNA re-sequencing (mapping, QC)SmallRNA-Seq (mapping, QC, gene count)Differential expression analysisAlternative splicing analysisGene ontology/ Pathway/ GSEA analysisVariant callingFusion genes detectionChip/CLIP-SeqTranscriptome assemblyCRISPR analysisCopy number analysisscRNA-primary analysismetagenomic 16S rRNA-SeqRNA-editing analysisFI MUNI Bioinformatics Seminar50
51. Thank youfor your attentionwww.ceitec.euCEITEC@CEITEC_Brno51