ITCR Collaboration Supplement Rachel Karchin, Mike Ryan, Shirley - - PowerPoint PPT Presentation

▶
itcr collaboration supplement
SMART_READER_LITE
LIVE PREVIEW

ITCR Collaboration Supplement Rachel Karchin, Mike Ryan, Shirley - - PowerPoint PPT Presentation

Sep 20, 2023 319 likes •833 views

ITCR Collaboration Supplement Rachel Karchin, Mike Ryan, Shirley Liu, Aviv Regev, Brian Haas, Mary Goldman, David Haussler, Jing Zhu UCSC Xena CRAVAT/MuPIT TCGA, ChIP-seq Annotated Xena Cistrome RNA-Seq LINCS, Data Transcriptome

slide-1
SLIDE 1

ITCR Collaboration Supplement

Rachel Karchin, Mike Ryan, Shirley Liu, Aviv Regev, Brian Haas, Mary Goldman, David Haussler, Jing Zhu

UCSC Xena CRAVAT/MuPIT

slide-2
SLIDE 2

Trinity Cistrome Xena Cravat/ MuPIT

RNA-Seq Data TCGA, LINCS, … ChIP-seq Data Browser Annotated Transcriptome Xena Browser Cistrome Analytics Mutations CRAVAT Annotation & Visualization

slide-3
SLIDE 3

Trinity Cistrome Xena Cravat/ MuPIT

RNA-Seq Data

Mutations Scores Annotations 3D Visualization

TCGA, LINCS, … ChIP-seq Data Browser Annotated Transcriptome Xena Browser Cistrome Analytics Mutations

TCGA Data

CRAVAT Annotation & Visualization

Mutations CRAVAT Driver Scores 3D Visualization

slide-4
SLIDE 4

Trinity Cistrome Xena Cravat/ MuPIT

RNA-Seq Data

Mutations Scores Annotations 3D Visualization

TCGA, LINCS, … ChIP-seq Data Browser Annotated Transcriptome Xena Browser Cistrome Analytics Mutations

TCGA Data

CRAVAT Annotation & Visualization

Mutations CRAVAT Driver Scores 3D Visualization

slide-5
SLIDE 5

CRAVAT/MuPIT

http://cravat.us/ High-throughput services to annotate and prioritize variants and genes.

Parallel Processing Server Annotations Features Mappings Algorithms

slide-6
SLIDE 6

CRAVAT/MuPIT

Funnel can be tailored to user needs

slide-7
SLIDE 7

Example of funnel developed for familial pancreatic cancer (FPC) WGS study

  • Genetic basis of FPC unknown for 80-90% of patients
  • Analysis of 70M SNVs identifies new genes not previously

implicated in FPC susceptibility

well-known FPC genes

slide-8
SLIDE 8

select from 31 TCGA cancers, mutations from UCSC Xena

CRAVAT/MuPIT interactive variant view

slide-9
SLIDE 9

CRAVAT/MuPIT interactive gene view

slide-10
SLIDE 10

CRAVAT / MuPIT ITCR Integration Supplement

  • Discussions with other ITCR tools showed need

for multiple ways of integrating with CRAVAT/MuPIT to maximize use of our services:

– Fast method of getting score and annotations for a single genomic location. – Asynchronous processing of full batch of input mutations. – Galaxy support needed – Simple method for other web tools to show MuPIT visualizations of their mutation in annotated 3D structures. – Cross platform support needed – request from other web servers, Java, Python, etc.

slide-11
SLIDE 11

Developed - CRAVAT Web Service

Annotations CRAVAT / MuPIT Web Server Program / Web Server DB SNP 1000 Genome ESP6500 ExAc COSMIC Dark Regions Precomputed VEST Scores (Mutation Pathogenicity) HTTP Request (genomic location) JSON Reply with: VEST Score Mutation Impact Cancer Annotations Population Stats Mapability Warnings

Fast, Cross-platform single mutation lookup

Driver Genes

slide-12
SLIDE 12

Developed - CRAVAT RESTful Service

Annotations CRAVAT / MuPIT Web Server Program / Pipeline DB SNP 1000 Genome ESP6500 ExAc COSMIC Dark Regions HTTP Post Request (file of mutations)

Method for automated submission of large sets of mutations

Driver Genes HTTP Status Check Results File

  • CHASM and VEST Scores
  • Full set of annotations

CRAVAT / VEST Algorithms

SNVBox Precomputed Genomic Features

slide-13
SLIDE 13

Developed MuPIT HTTP Link Integration

MuPIT / PDB Database CRAVAT / MuPIT Web Serverr External Web Page HTTP URL HTML / Javascript Visualization Page

Get MuPIT 3D Mutation Visualization with URL with genomic location or gene

slide-14
SLIDE 14

Developed Galaxy Integration

CRAVAT tool in Galaxy Toolshed – VEST Scores and Annotations for Mutations

slide-15
SLIDE 15

Trinity Cistrome Xena Cravat/ MuPIT

RNA-Seq Data

Mutations Scores Annotations 3D Visualization

TCGA, LINCS, … ChIP-seq Data Browser Annotated Transcriptome Xena Browser Cistrome Analytics Mutations

TCGA Data

CRAVAT Annotation & Visualization

Mutations CRAVAT Driver Scores 3D Visualization

slide-16
SLIDE 16

Trinity CTAT Mutation Pipeline Collaborations Overview

slide-17
SLIDE 17

Access to supercomputing resources from your laptop!

Trinity CTAT Mutation Pipeline Collaborations Overview

slide-18
SLIDE 18

Transparent integration to CRAVAT web services.

Trinity CTAT Mutation Pipeline Collaborations Overview

slide-19
SLIDE 19

Custom visualization is a product of multiple labs efforts.

Trinity CTAT Mutation Pipeline Collaborations Overview

slide-20
SLIDE 20

Sample Visualization is a Collection of Collaborations

slide-21
SLIDE 21

Sample Visualization is a Collection of Collaborations IGV

slide-22
SLIDE 22

Sample Visualization is a Collection of Collaborations CRAVAT 

slide-23
SLIDE 23

Sample Visualization is a Collection of Collaborations NCGAS

slide-24
SLIDE 24

 Sample Visualization is a Collection of Collaborations

slide-25
SLIDE 25

Exploring a Single Variant

slide-26
SLIDE 26

Exploring a Single Variant IGV

slide-27
SLIDE 27

Exploring a Single Variant CRAVAT

slide-28
SLIDE 28

Exploring a Single Variant MuPIT

slide-29
SLIDE 29

 Exploring a Single Variant

slide-30
SLIDE 30

MuPIT Provides Additional Context

Visualize variants within protein structure.

slide-31
SLIDE 31

Trinity Cistrome Xena Cravat/ MuPIT

RNA-Seq Data

Mutations Scores Annotations 3D Visualization

TCGA, LINCS, … ChIP-seq Data Browser Annotated Transcriptome Xena Browser Cistrome Analytics Mutations

TCGA Data

CRAVAT Annotation & Visualization

Mutations CRAVAT Driver Scores 3D Visualization

slide-32
SLIDE 32

Cistrome Analysis Pipeline Functions

  • ChIP-chip / seq peak calling
  • Checking corr and overlap
  • Visualize signal across different elements
  • Conservation analysis
  • Lift over from one genome to another
  • Annotate target genes
  • Motif analysis
  • Heatmap and clustering of many factors
  • Gene expression microarray analysis
  • Examine factor function and collaborators
  • Compatible with Genome Space

Liu et al, Genome Biol 2011; > 6000 users, 64TB of data

32

slide-33
SLIDE 33

33

Cistrome Dataset Browser

slide-34
SLIDE 34

RABIT Integration framework

Key features: 1, Control tumor confounding factors (CNA, Methylation) 2, Test ENCODE cell line match with TCGA cancer types

RABIT framework

slide-35
SLIDE 35

collaborate RABIT XENA: visualization

slide-36
SLIDE 36

RABIT -> Xena <- Cistrome

Samples n-500

We host the RABIT results

  • n main UCSC public hub

along with all other Level 2+ TCGA data, such as gene expression.

RABIT SPI1 regulatory impact SPI1 From RABIT From TCGA

RABIT t-test value indicates the regulatory potential of a specific transcription factor (TF) in a tumor

TF up-regulates its target genes TF down-regulates its target genes TF does not regulate its target genes Cistrome Identified SPI1 ChipSeq Target SPI1 target genes

slide-37
SLIDE 37

p-value = 0.00022 Samples RABIT SPI1 regulatory impact

Tumors where SPI1 down-regulates its target genes (low score) are associated with poor survival in TCGA Kidney Clear Cell Carcinoma (KIRC).1

[1] Jiang P, et. al.. Inference

  • f transcriptional regulation

in cancers. Proc Natl Acad Sci U S A. 2015 Jun 23; 112(25):7731-6.

RABIT -> Xena

slide-38
SLIDE 38

Linc-RNA super enhancer curation from histone mark ChIP-seq

slide-39
SLIDE 39

Linc-RNA promoter annotation and Trinity collaboration

Waiting for linc-RNA analysis result from Trinity

slide-40
SLIDE 40

Trinity Cistrome Xena Cravat/ MuPIT

RNA-Seq Data

Mutations Scores Annotations 3D Visualization

TCGA, LINCS, … ChIP-seq Data Browser Annotated Transcriptome Xena Browser Cistrome Analytics Mutations

TCGA Data

CRAVAT Annotation & Visualization

Mutations CRAVAT Driver Scores 3D Visualization

slide-41
SLIDE 41

Your private data

UCSC Xena

TCGA ICGC CCLE

Treehouse Childhood Cancer

UCSC Xena System

Xena Hub

Hosts functional genomics data. API access. Easily deployable: MAC, Windows, Linux, Amazon cloud

Xena Browser

Web-based data access and visualization across multiple Xena hubs.

Integrated with Galaxy

Connect to many other bioinformatics tools, including mutation callers, clustering algorithms, etc.

xena.ucsc.edu

slide-42
SLIDE 42

From main, public Xena Hub Every row is a sample, 467 samples From private Xena Hub on laptop

Visualize data from two Xena Hubs on a web browser

slide-43
SLIDE 43

UCSC Xena Main Data Hub

  • ICGC (release 19)

–42 cohorts, plus ICGC Pan-Cancer data

  • TCGA (level 2+)

–34 cancer types, plus TCGA Pan-Cancer data

  • Pediatric Cancer

–St. Jude PCGP, TARGET, Treehouse etc

  • Cell Line data
slide-44
SLIDE 44

Supported data types

  • Mutation: SNPs and small INDELs
  • Gene, Exon, miRNA, Protein expression (-seq
  • r Array-based)
  • Copy Number
  • DNA Methylation
  • Phenotype/Clinical data: e.g. subtype, tumor

type, age

slide-45
SLIDE 45

Somatic Mutation Analysis

CRAVAT -> Xena -> MuPIT

slide-46
SLIDE 46

CRAVAT -> Xena -> MuPIT

Samples (n=437)

Xena gene-level somatic mutation heatmap:

Top 10 mutated genes (MutSig) in TCGA Bladder Cancer Cohort

Samples (n=437)

RHOB TP53 RB1 ELF3 TSC1 PIK3CA KDM6A CDKN2A ARID1A STAG2

Gene harbors any non-silent mutation Gene does not harbor a non-silent mutation

slide-47
SLIDE 47

CRAVAT -> Xena -> MuPIT

RHOB TP53 RB1 ELF3 TSC1 PIK3CA KDM6A CDKN2A ARID1A STAG2

Xena displays RHOB mutations

  • n transcript.

Transcript model along top. Mutations ordered by position on the transcript. CRAVAT driverness score added by user.

Samples

VAF 100% VAF 0% q value 0 q value 0.5

RHOB mutation location RHOB variant allele frequency CRAVAT driverness score Samples

slide-48
SLIDE 48

VAF 100% VAF 0% q value 0 q value 0.5

RHOB mutation location RHOB variant allele frequency CRAVAT driverness score

CRAVAT -> Xena -> MuPIT

Single click transfers mutations in Xena to MuPIT 3D- view

Samples

slide-49
SLIDE 49

driverness on structure 1cxz (MuPIT)

CRAVAT -> Xena -> MuPIT

Mutations with the most significant CHASM q values are at the interface with GTP and effector protein PKN/RPK1

q value 0 q value 0.5

CRAVAT driverness score

AA with mutations with high driverness score GTP Effector domain of protein kinase PKN/PRK1 Crystal structure of RHOA GTPase (RHOA and

RHOB share 85% amino acid sequence identity)

slide-50
SLIDE 50

Acknowledgments

U01CA180956 U01CA180980 U24CA180922 U24CA180951 ITCR Administrative Supplement