MS523.L18.Clincial Genomics.Q3.23.pptx
Document Details
Uploaded by LogicalSquirrel
Full Transcript
Lecture: Clinical Genomics Presenter: Dr. Darl Swartz Course: Human Genetics MS523 Date: 3/3/23 12/12/ Dr. Darl Ray. Swartz 1 Objectives: 1. Describe the evolution of DNA sequence analysis in clinical diagnostics and application. 2. Explain why comparing a patient’s genome to a reference genome...
Lecture: Clinical Genomics Presenter: Dr. Darl Swartz Course: Human Genetics MS523 Date: 3/3/23 12/12/ Dr. Darl Ray. Swartz 1 Objectives: 1. Describe the evolution of DNA sequence analysis in clinical diagnostics and application. 2. Explain why comparing a patient’s genome to a reference genome does not readily inform you of the potential causative gene. 3. Explain how cascade/trio whole genome sequencing more readily allows for variant calling than comparison of a proband’s genome to a reference genome. 4. Describe Geisinger’s MyCode goals and focus areas. 5. Explain how nationalized medicine and genomics can result in personalized medicine from cradle to grave. 6. Explain the challenges of undiagnosed diseases and how genomics and AI can facilitate diagnosis. 7. Describe some the changes that occur in the human genome (mitochondrial and nuclear) as one ages (>70 yo) 8. Describe how trio WGS can be used to determine recombination hot spots and de-novo mutations 12/12/2 Dr. Darl Ray. Swartz 2 Outline: I. II. Linking Genomes to Phenomes Geisinger’s Genomic Medicine Institute III. ClinGen and Other Consortiums IV. Undiagnosed Diseases V. Trio and Zygotic Twin Comparative Genomics 12/12/2 Dr. Darl Ray. Swartz 3 inking Genomes to Phenomes 12/12/2 Dr. Darl Ray. Swartz 4 inking Genomes to Phenomes A) Genome Analytical/Diagnostic Methods 1) Multiplex PCR panels for a specific disease (a) Determine which variant the proband has 2) Custom microarrays allowing actionable results for genetic counseling or treatment (a) Monogenic diseases with specific high frequency allele (b) Disease risk genes from GWAS or linkage analysis (c) Polygenic genes with high-risk scores (d) Drug metabolizing enzymes (e) Drug target variants 12/12/2 Dr. Darl Ray. Swartz 3) Targeted sequencing using phenomic 5 inking Genomes to Phenomes A) Genome Analytical Methods (ca $1,000/person now) 4) Whole genome (WGS) and exome (WES) sequencing (a) Genome includes coding and noncoding regions (b) Exome includes just known coding genes and a few bases into the introns (i) Fragment DNA (ii) Ligate universal primer region (iii) Enrich for exons using hybridization Array or in solution (iv) Amplify via universal primer (v) Downstream processing for NGS (vi) Detects coding and splice variants (c) Compare sequence to “standard” 12/12/2 Dr. Darl Ray. Swartz (i) Problem of what reference genome to 6 inking Genomes to Phenomes A) Genome Analytical Methods 5) Cascade (family)/trio (and higher) whole genome or exome sequencing (a) Whole genome sequencing of all family members of the proband (patient) (i) Within a family generation Need parents and sibs (ii) 2 – 3 generations even better (b) Compare sequences amongst family members and to reference genome (i) Intrafamily comparison dramatically reduces the offtarget differences (c) Akin to a single base pair 12/12/2 resolution linkage study Dr. Darl Ray. Swartz 7 inking Genomes to Phenomes A) Genome Analytical Methods 5) Cascade (family) whole genome or exome sequencing (e) Allows for: (i) Identification of inheritance and recombination sites (ii) More rapidly identify candidate genes based upon potential mode of transmittance Dominant or recessive at very low frequency Compound heterozygotic inheritance (iii) Identify de novo gametic and somatic mutations 6) Current challenge for insurance coverage 12/12/2 Dr. Darl Ray. Swartz in children and adults because of coding 8 inking Genomes to Phenomes B) Phenomic methods 1) Digitized patient data (a) Structures: Morphology, MRI, Xray, Pathology slides (b) Metabolomic data: Blood, Urine, Sweat, Salivary (c) Physiological parameters: ECG, EMG, Respiratory parameters, Blood pressure, Glomerular filtration rate 2) Patient interview/history (a) Current challenge to digitize/standardize terms 3) Need for consolidation of data in digital format initiated by Dept. of Health and Human Services for electronic health data > electronic health records 12/12/2 Dr. Darl Ray. Swartz 9 inking Genomes to Phenomes B) Phenomic methods 4) Most hospitals now use EHR for data entry and analysis for treatment (a) Doctors spend most of their time entering data instead of seeing patients 5) Can share EHR data within and between hospitals within and outside of networks allowing for rapid communication 6) Linking phenomic data from EHRs and genomic data allows for data mining (a) AI based learning and diagnosis (b) Bigger/more data means better AI 7) Requires large, secure, technology infrastructure 8) Current work focuses on linking the data > eMerge 12/12/2 Dr. Darl Ray. Swartz 10 inking Genomes to Phenomes C) Clinical utility of genomic sequencing considerations: 1) Technical efficacy > sequencing capabilities 2) Diagnostic accuracy efficacy > variant classification 3) Diagnostic thinking efficacy > help in diagnosis along with phenomic data 4) Therapeutic efficacy > help in treatment decisions 5) Patient outcome efficacy > improve patient outcome 6) Societal efficacy > is it cost effective npj Genomic Medicine (2020)5:56 ; https://doi.org/10.1038/s41525-020-00164-7 12/12/2 Dr. Darl Ray. Swartz 11 inking Genomes to Phenomes D) Current challenges in the use of EHR for genomic medicine via “Genomic Information for Clinicians in the Electronic Health Record: Lessons Learned From the Clinical Genome Resource Project and the Electronic Medical Records and Genomics Network” 2019 Frontiers in Genetics 10:1059 1) Lack of standards to represent and communicate genomic information 2) Inability to store genomic information in current EHR systems 3) Translating genomic variants into clinical phenotypes that clinicians can recognize and use to manage patients 4) Access to reliable genomic knowledge sources 5) Existing efforts are largely supported by institutional and grant funding. Sustainable models are needed for further development 12/12/2 Dr. Darl Ray. Swartz 12 inking Genomes to Phenomes E) Current status of integrating genomic medicine into primary care in Canada (via questionnaires – Fronters in Genetics 2019, 10:1189) 12/12/2 Dr. Darl Ray. Swartz 13 inking Genomes to Phenomes E) Current status of integrating genomic medicine into primary care in Canada (via questionnaires – Fronters in Genetics 2019, 10:1189) 12/12/2 Dr. Darl Ray. Swartz 14 inking Genomes to Phenomes E) Current status of integrating genomic medicine into primary care in Canada (via questionnaires – Fronters in Genetics 2019, 10:1189) 1) Conclusions from family physician questionnaires: (a) See their role in developing family history/pedigrees, making referrals, and supporting patients with genetic diseases (b) Lack confidence in genomic medicine skills needed for practice (c) Somewhat optimistic about contributions of genomic medicine but are cautious about its current clinical benefits 2) Needs are: (a) Training in obtaining a family history/pedigree (b) Evidence of clinical utility of genetic tests (c) Educational resources that can be integrated into primary care practice, clinical decision support, and improved communication with genetic specialists 3) Educational approach ? Case studies starting with the first presentation of the patient in the primary care setting – like you have been doing!!! 12/12/2 Dr. Darl Ray. Swartz 15 eisinger’s Genomic Medicine Institute 12/12/2 Dr. Darl Ray. Swartz 16 eisinger’s Genomic Medicine Institute A) Data integrator and generator of genomic and EHR for precision medicine via development of eMERGE B) Developed ClinGen site containing clinical relevance of genes and variants (genotype > disease phenotype) 12/12/2 Dr. Darl Ray. Swartz 17 eisinger’s Genomic Medicine Institute C) Started MyCode for precision medicine initiative at Geisinger Hospitals 1) Use custom microarray to screen for various disease risk alleles or exome sequencing in collaboration with Regeneron 2) Focus is on actionable disease risk (a) Phenomic testing for confirmation and subsequent treatment (b) Increased testing frequency for cancer risk alleles (c) Inform family members (d) Life-style changes 12/12/2 Dr. Darl Ray. Swartz 18 eisinger’s Genomic Medicine Institute D) Main risk conditions studied 1) CDC tier 1 conditions (a) Hereditary breast and ovarian cancer via BRCA1 and BRCA2 (b) Lynch syndrome (early colon and uterine cancers) via PMS2, MSH6, MSH2, MLH1 (c) Familial hypercholesteroemia (early heart attack and strokes) via APOB, LDLR 2) Cardiovascular risk (a) Cardiomyopathy (hypertrophy or dilated myopathy) via many sarcomeric proteins genes (b) Arrhythmias via channel protein gene variants (c) Arrhythmogenic right ventricular cardiomyopathy via desmosome associated proteins (d) Marfan Syndrome via FBN1 19 12/12/2 Dr. Darl Ray. Swartz (e) Heritable thoracic aortic disease (aortic aneurism) via smooth eisinger’s Genomic Medicine Institute D) Main risk conditions studied 3) Cancer (a) Hereditary pheochromocytomas and paragangliomas via succinate dehydrogenase genes (b) Multiple endocrine neoplasia type 1 via MEN1 tumor suppressor gene (c) Multiple endocrine neoplasia type 2 via RET proto-oncogene (d) PTEN hamartoma tumor syndrome via PTEN variants (e) Tuberous sclerosis via TSC1 and TSC2 gene (f) Li-Fraumeni syndrome via TP53 gene (g) Familial adenomatous polyposis via APC gene (h) Von Hippel-Lindau syndrome via VHL gene (i) Retinoblastoma via RB1 gene 12/12/2 Dr. Darl Ray. Swartz 20 eisinger’s Genomic Medicine Institute D) Main risk conditions studied 4) Other (a) Malignant hyperthermia via RYR1 gene (b) Fabry disease via GLA gene (c) Vascular Ehlers-Danlos Syndrome via COL3A1 gene (d) Hereditary hemochromatosis via HFE gene (e) Hereditary hemorrhagic telangiectasia via ENG gene (f) Juvenile polyposis syndrome via SMAD4 12/12/2 Dr. Darl Ray. Swartz 21 eisinger’s Genomic Medicine Institute E) To date (2/1/2023) have over 300,000 participants resulting in: 12/12/2 Dr. Darl Ray. Swartz 22 eisinger’s Genomic Medicine Institute F) Goal is 2 million patient data sets for analysis G) Collaborated with Regeneron to analyze the samples via microarrays 1) Regeneron has access to de-identified genomic and phenomic data for drug development and variant discovery H) Regeneron organized a consortium with others in Big Pharma to fund $50M project to sequence UK Biobank samples 1) UK Biobank similar to Geisinger program but uses NHS phenomic data and has 500,000 samples 2) Can exome sequence up to 200,000 patients a year 3) Consortium had exclusive access to data for one year 4) Will allow for more refined GWAS and clinically actionable variants 5) Can identify targets for drug development 6) Can allow for refined clinical trials on targets 12/12/2 Dr. Darl Ray. Swartz 23 eisinger’s Genomic Medicine Institute I) Current challenges are noted by the American College of Medical Genetics and Genomics (The interface of genomic information with the electronic health record: points to consider statement of the American College of Medical Genetics and Genomics (ACMG). Genet. Med. 22, 1431–1436 (2020).) 1) Incorporation into electronic health record 2) Patient accessibility 3) Inter-institution/network accessibility of data 4) Linking genomic data to clinical decision and management support tools 5) Coding issues for un-common genetic diseases (coded as a chromosomal aberration) 6) Patient data protection when used for disease 12/12/2 Dr. Darl Ray. Swartz npj Genomic Medicine (2019)4:12 ; https://doi.org/10.1038/s41525-019-0085-8 24 ClinGen and Other Consortiums 12/12 25 ClinGen and Other Consortiums A) US and other developed countries are aggregating clinical phenomic and genomic data to mine for disease diagnosis B) USA has ClinGen 1) A National Institutes of Health (NIH)-funded resource dedicated to building a central resource that defines the clinical relevance of genes and variants for use in precision medicine and research 2) Focuses on (a) Gene-Disease validity > Can variation in this gene cause disease? (b) Variant pathogenicity > Which changes in the gene cause the disease (c)Dosage sensitivity > Does loss or gain of this gene or genomic region result in disease (d) Somatic cancer variants > Clinical significance of genomic anomalies associated with different cancer types (e) Baseline annotation > Annotation of the genome based upon the biomedical literature 26 12/12 (f) Clinical actionability > Are there actions that could be taken to ClinGen and Other Consortiums B) USA has ClinGen 3) Gene-Disease Validity tool has 1619 curated genes 4) Dosage Sensitivity tool has 1996 curated dosage regions 5) Variant Pathogenicity tool 6) Actionability knowledge repository tool 12/12 27 ClinGen and Other Consortiums C) Many other countries with nationalized medicine are harvesting phenomicgenomic relationships 1) UK BioBank 2) Several Nordic countries with family clinical histories of three generation or more 3) Efforts underway to combine many D) Medical Gemone Reference Bank contains whole genome and phenome data for 2,570 healthy elderly > what happens to the genome as you get old? 1) Age > 70 2) Used blood as the source of DNA 3) Analyzed several genomic and a few physiological features (a)Telomere length > no change (b)Mitochondrial number > decline (c) Y-copy number in males > decline (d)Somatic variant burden > increase (e)Mitochondrial variants > increase (f) Clonal haemotopoiesis > increase 12/12 28 ClinGen and Other Consortiums D) Medical Gemone Reference Bank contains whole genome and phenome data for 2,570 healthy elderly > what happens to the genome as you get old NATURE COMMUNICATIONS | (2020)11:435 | https://doi.org/10.1038/s41467-01914079-0 | www.nature.com/naturecommunications 12/12 29 Undiagnosed Diseases 12/12 Dr. Darl Ray. Swartz 30 Undiagnosed Diseases A) Diseases in which current phenotypic or genotypic knowledge does not diagnose the disease 1) Rare variants in the population with no documented (or discovered) record of diagnosis 2) Estimates of 7,000 diseases 3) Usually takes 7.6 years to diagnose in the US 4) See numerous primary care and specialist and difficulty coordinating information between clinicians 12/12/ Dr. Darl Ray. Swartz 31 Undiagnosed Diseases B) Genetics of disease is difficult to determine 1) Very low frequency in population for mono-genetic diseases 2) Complex heterozygote (a) Two recessive alleles at the same locus with one on each chromosome 3) De novo mutations 4) Polygenic diseases C) Many consortia established to share phenomic and genomic data intra and intercontinental 1) Gives bigger data base for comparing like phenomes/genomes D) More objective phenotypic data to facilitate diagnosis using human phenotype ontology 1) Develop structured phenotypic data for use in various computational 12/12/ Dr. Darl Ray. Swartz applications 32 Undiagnosed Diseases Marwaha et al. Genome Medicine (2022) 14: E) Genomics applied to determine gene(s) involved in diagnosis going from low to SNP resolution 1) 2) 3) 4) 5) 6) 7) 8) 9) 12/12/ Karyotyping Chromosomal microarray Disease microarrays Exome and whole exome sequencing Genome sequencing Transcriptomics Epigenomics Proteomics Sequencing efforts using parent-fetus/offspring trios (a) Like cascade sequencing to quickly rule out non- Dr. Darl Ray. Swartz 33 Undiagnosed Diseases F) Diagnosis can be facilitated by adding in other biological knowledge from cell and animal models 1) Genetic manipulation of cell and animal models to observe phenomic effects G) Using these combinatorial methods has improved diagnosis by about 10% to date over the 25 – 35% diagnostic rate w/o leveraging recourses 1) Diagnostic rates decrease with increasing age of presentation (a) Potential environmental components 12/12/ Darl Ray. Swartz Marwaha et al. Dr. Genome Medicine (2022) 14:23 34 Undiagnosed Diseases H) Current efforts leverage artificial intelligence by collecting data from numerous sources and analyzing it 1) Computer programs trained on existing data (a)Deep-learning and neural network based (b) Synthesizes phenomic and genomic data (c)Leverages published and repository data (d) Used to develop and improve Face2Gene 2) Uses data from unknown to predict genetic cause (a)i.e. what a well-trained clinical 12/12/ Dr. Darl Ray. Swartz 35 io and Zygotic Twin Comparative Genomics Halldorsson et al., Science 363, eaau1043 (2019) 25 January 2019 NATURE GENETIcS | VOL 53 | JANUARy 2021 | 27–34 12/12/ Dr. Darl Ray. Swartz 36 io and Zygotic Twin Comparative Genomics A) Several GWAS and trio/twin studies done via deCODE, an Iceland company in collaboration with Amgen (a biotech company) 1) Leverages unique population genetics and social features 12/12/ (a)Low genomic diversity mostly from founder effects of island community (b) Socialized medicine and medical records (c)Extensive genealogical history of residents (d) Extensive population genomic data (both high resolution array and more recently WGS) Dr. Darl Ray. Swartz 37 io and Zygotic Twin Comparative Genomics B) Trio study on recombination frequency show several features 1) Trio of mother, father, and offspring allows for determination of crossover location, de novo mutations, type of mutation, sex differences, and age effects 2) Crossover occurs mostly (ca 80%) in the same locations called cross-over hotspots 3) Can have complex cross-over where there is both a cross-over and a gene conversion on the sister at relatively low frequencies (a)1.24% of crossovers for females (b) 0.54% of crossovers for males 12/12/ Dr. Darl Ray. Swartz 38 io and Zygotic Twin Comparative Genomics B) Trio study on recombination frequency show several features 4) Crossover number increases with maternal age but not paternal (a)Difference in gametogenesis, particularly meiosis 12/12/ Dr. Darl Ray. Swartz 39 io and Zygotic Twin Comparative Genomics B) Trio study on recombination frequency show several features 5) Crossover results in de novo mutations (a)Average of about 70 per trio (206,000/3000 trios) (i) About 4 X 10-7 per base per generation or 50X that in other locations (b) Most mutations were SNPs (i) Type of SNP (transition or transversion) differed between males and females likely related to chromatin state differences between female and male gametogenesis Male DNA methylated 12/12/ Dr. Darl Ray. Swartz 40 io and Zygotic Twin Comparative Genomics B) Trio study on recombination frequency show several features 6) De novo mutations are (likely evolved to) in non-coding regions with most in DNA regulatory elements (a)Potentially modify transcription of nearby coding genes 7) Most de novo mutations within 1kb of crossover site (a)Excise region then repair via a bit sloppy DNA polymerases (not sure as to which polymerases are involved at present) 8) Did GWAS for recombination rate and associated metrics and found several hits of SNP variants in meiosis related proteins that altered function or transcription/translational regulation (a)Recombination rates have a Dr. genetic 12/12/ Darl Ray. Swartz 41 io and Zygotic Twin Comparative Genomics C) deCODE has also done monozygotic twin GWAS studies to compare how similar they are and if dissimilar, when did the de novo mutation occur 1) Observed a median of 15 – 48 (low coverage – high coverage reads) postzygotic de-novo mutations (discordant) in 381 twin pairs (a)One twin pair with 25,653 discordant de novo mutations 2) Observed an increase in post-zygotic mutations with increasing age (a)More observed in DNA from blood than oral mucosa (i) Clonal selection of blood cells with increasing age 12/12/ Dr. Darl Ray. Swartz 42 Copyright Notice All materials found on Geisinger Commonwealth School of Medicine’s course and project sites may be subject to copyright protection, and may be restricted from further dissemination, retention or copying. Disclosure I have no financial relationship with a commercial entity producing health-care related products and/or services.