Welcome aboard!

Welcome to CCI wiki! Here you'll find everything you need to know about CCI.

You can read more about the Center for Molecular Oncology (CMO) here:

MSK-ACCESS (Research)

Developed by scientists in the CMO Technology Innovation Lab and Department of Pathology, this high-sensitivity assay is offered by the CMO to MSK researchers for profiling circulating tumor DNA derived from blood plasma. The inclusion of matched buffy coat DNA enables the identification and elimination of germline variants and mutations associated with clonal hematopoiesis, a significant confounder of most commercial assays. The assay is available for clinical use in the Molecular Diagnostics Service and for research projects in the Integrated Genomics Operations (IGO). CCI supports the data processing and analysis of research projects utilizing MSK-ACCESS in IGO and leads the ongoing development of the MSK-ACCESS pipeline for all applications. The current version of the pipeline is available here: mskcc/ACCESS-Pipeline: cfDNA Sequencing Pipeline with UMI (github.com), and more details about the assay and analysis are described in this paper below as well as here:

Brannon, A. R. et al. Enhanced specificity of clinical high-sensitivity tumor mutation profiling in cell- free DNA via paired normal sequencing using MSK-ACCESS. Nat Commun 12, 3770 (2021).

CMO-CH

Developed by scientists in the CMO Technology Innovation Lab in collaboration with CCI and Clonal Hematopoiesis (CH) program, Diagnostic Molecular Pathology, Precision Interception, and Prevention Initiative & CCI, this assay utilizes the same barcoding and ultra-deep sequencing technology as MSK-ACCESS to detect CH mutations in white blood cells at high sensitivity. CMO-CH is offered by the CMO to MSK researchers for profiling white blood cell DNA to detect mutations in the most commonly altered CH-associated genes. The assay is run in IGO, and CCI supports the data processing and analysis. You can learn more about it here

For both projects, additional analysis packages and development versions of the workflows can be found here: CMO Cell-Free DNA (cfDNA) Informatics Team (github.com)

What is CMO-CH assay?

CMO-CH is offered by the Center for Molecular Oncology (CMO) to MSK researchers for profiling white blood cell DNA to detect mutations in the most commonly altered clonal hematopoiesis (CH) associated genes.

What is the panel design?

• 596 targets capturing 58% of CH and 90.4% of CH-PD mutations identified in the latest CH dataset from 40K patients

• Total size = 1,143 probes (0.14 Mb)

• Full gene coverage for TP53, TET2, ASXL1, DNMT3A, PPM1D, CHEK2, ASXL1, ATM, SF3B1, SRSF2, U2AF1, and U2AF2•Additional targets with hotspot positions from IMPACT heme assay

• SNP tiling around TP53, CBL, MPL, JAK2, EZH2, TET2, RUNX1, and ATM (+/-10kb) to identify allelic-imbalances

• 40 fingerprint SNPs that are shared with all other NGS assays (IMPACT, ACCESS, WES etc.) to detect sample mismatches

Our Mission

The CMO Cell-Free DNA Informatics (CCI) group’s mission is to develop and apply computational methods to organize, analyze and understand genomic data generated from cfDNA assays such as MSK-ACCESS. The group is responsible for all computational infrastructure needed to deploy, run, and deliver results for research cfDNA assays in a production setting for CMO.

Our Values

Be Compassionate

We treat everyone we encounter with compassion, seeing the humanity behind their problems and experiences.

Be Mindful

We do not take advantage of our users' attention and adopt mindful working practices so that we can create safe spaces both in our working environment and in our products themselves.

Research First

We challenge our own and others' assumptions through qualitative and quantitative research. Not sure about an idea? Test it.

Ronak H Shah

👋 Lead Scientist — 💌 shahr2@mskcc.org — 🇺🇸 New York

Bio

I am responsible for leading a team of Computational Biologists and Bioinformatics Software Engineers who develop, maintain, and operate bioinformatics pipelines and databases in the Center for Molecular Oncology. We also perform collaborative research with other labs and clinicians both within MSKCC and in the broader research community. On a daily basis, we analyze blood samples from patients with tissue-based cancers, using patients’ circulating tumor DNA extracted from blood, avoiding the need for tumor biopsies. More specifically, I lead the team in designing, developing, and implementing software tools for processing and analyzing high throughput, next-generation sequencing data, specifically for liquid biopsy applications (MSK-ACCESS & Clonal Hematopoiesis Panel). Previously in my time as MSK, I helped develop the workflow to analyze MSK-IMACT data for both clinical and research implementation, which is still being used. You can read more about my background here, also here is the link to my google scholar profile

Karthigayini Sivaprakasam

👋 Senior Computational Biologist — 💌 sivaprk@mskcc.org — 🇺🇸 Texas

Bio

Carmelina Charalambous

👋 Computational Biologist — 💌 charalk@mskcc.org — 🇺🇸 New York

Bio

I am a Computational Biologist as part of the Center of Molecular Oncology Informatics cfDNA group. I am responsible for the development of new and existing tools, maintaining the cfDNA pipeline, and processing samples for both clinical and research work. Previously, I have worked at the NHS trust Cambridge University Hospitals as a Bioinformatician with primary responsibility for the development of the Hemato-Oncology clinical assay. You can read more about my educational background, work experience, and research interests here and you can access my Linkedin profile.

Eric Buehler

👋 Bioinformatics Engineer II — 💌 buehlere@mskcc.org — 🇺🇸 New York

Bio

Alyssa Vann

👋 Bioinformatics Engineer IV — 💌 vanna1@mskcc.org — 🇺🇸 New York

Bio

BAM Generation & Quality Control

Github Location -> https://github.com/mskcc/ACCESS-Pipeline/blob/master/workflows/ACCESS_pipeline.cwl

Tools Used:

• BWA

• Trimgalore

• GATK

• Picard Tools

• ABRA2

• Marianas

Variant Calling

Small Variants

Github Location -> https://github.com/mskcc/ACCESS-Pipeline/blob/master/workflows/subworkflows/snps_and_indels.cwl

Tools Used:

• VardictJava

• MuTect

• VCF2MAF

Copy Number Variant (CNV)

Github Location -> https://github.com/mskcc/ACCESS-Pipeline/blob/master/workflows/subworkflows/call_cnv.cwl

Refer to Bioinformatics Pipeline to Detect CNA's section in this paper for details:

Dara S. Ross, Ahmet Zehir, Donavan T. Cheng, Ryma Benayed, Khedoudja Nafa, Jaclyn F. Hechtman, Yelena Y. Janjigian, Britta Weigelt, Pedram Razavi, David M. Hyman, José Baselga, Michael F. Berger, Marc Ladanyi, Maria E. Arcila, Next-Generation Assessment of Human Epidermal Growth Factor Receptor 2 (ERBB2) Amplification Status: Clinical Validation in the Context of a Hybrid Capture-Based, Comprehensive Solid Tumor Genomic Profiling Assay, The Journal of Molecular Diagnostics, Volume 19, Issue 2, 2017, Pages 244-254, ISSN 1525-1578, https://doi.org/10.1016/j.jmoldx.2016.09.010.

Structural Variant (SV)

Github Location -> https://github.com/mskcc/ACCESS-Pipeline/blob/master/workflows/subworkflows/manta.cwl

Tool Used:

• Manta

• iAnnotateSV

Microsatellite Instability Status (MSI)

Github Location -> https://github.com/mskcc/ACCESS-Pipeline/blob/master/workflows/subworkflows/msi.cwl

Tool Used:

• ADMIE

Configurations

Voyager has all our configurations in the jinja template, it includes all the paths for various files and tools associated with the workflows, all location are on JUNO:

snps_indels: beagle/input_template.json.jinja2 at master · mskcc/beagle (github.com)

CNV: beagle/input_template.json.jinja2 at master · mskcc/beagle (github.com)

Fastq_to_bam: beagle/input_template.json.jinja2 at master · mskcc/beagle (github.com)

MSI: beagle/input_template.json.jinja2 at master · mskcc/beagle (github.com)

SV : beagle/input_template.json.jinja2 at master · mskcc/beagle (github.com)

• The best place to start would be to understand the data flow:

  • LIMS/SMILE -> Voyager -> JUNO

• Learn more about the codebases listed below that form Voyager

• Learn about the codebases the form mPATH

Sites

Code/Project Management

Applications

People

Projects, who to speak with, slack channel

• MPath & CVR (Aijazuddin Syed/Anoop Balakrishnan Rema)

• Voyager Projects [Beagle/Seqosystem/Ridgeback] (Sinisa Ivkovic/Nikhil Kumar/Allan Bolipata)

  • #voyager

• LIMS (David Mcmanamon)

• MDB (Angelica Ochoa/Benjamin Gross/Allan Bolipata)

  • #metadb-informatics

• Toil/CWL (Nikhil Kumar)

• ACCESS (Ronak Shah)

  • #msk-access

• ACCESS Servers (HPC Request/Neeraj Paramasivam)

CMO onboarding

The first thing to do once you have joined is to visit https://mskcc.github.io/on-boarding/ and finish of task necessary for compliance & initiate the process of getting access to various systems.

Cluster Guide

To learn more about the cluster and its resources visit the MIRO board

CCI specific onboarding

Guides on Confluence

• The best place to start is to learn more about MSK-ACCESS and for that please read the paper:

  • Brannon, A. R. et al. Enhanced specificity of clinical high-sensitivity tumor mutation profiling in cell- free DNA via paired normal sequencing using MSK-ACCESS. Nat Commun 12, 3770 (2021).

• Learn more about the current collapsing method Marianas

• Learn more about the new collapsing (Nucleo) method using Fgbio

• Learn more about the Quality Control V1

• Understand the updated version for the above using these Quality Control V2

• Learn about ACCESS Data analysis scripts that help with downstream analysis

• Learn about IGV for viewing BAM files to distinguish real variants from artifacts

Project Management

Functional Resources

Path's to know

CMO-ACCESS

Data on JUNO for CMO-ACCESS samples

Resources

Details of CMO ACCESS Resources

• admie - Files used for microsatellite instability detection tool ADMIE for MSK-ACCESS

• cosmic - VCF file of cosmic used in MSK-ACCESS workflows

• dbSNP - VCF file of dbSNP used in MSK-ACCESS workflows

• exac - VCF file of ExAC used in MSK-ACCESS workflows

• mills-and-1000g - VCF file of mills-and-1000g used in MSK-ACCESS

• reference - reference genome file used in MSK-ACCESS workflows

• tools - general packages used in MSK-ACCESS workflows

• msk-access - Data-specific resources for MSK-ACCESS workflows. This includes the following:

  • hiseq4000_curated_duplex_bams_dmp - curated DMP duplex BAMS from HiSeq 4000

  • novaseq_curated_simplex_bams_dmp - curated DMP simplex BAM from NovaSeq.

  • hiseq4000_curated_simplex_bams_dmp - curated DMP simplex BAM from HiSeq 4000

  • novaseq_curated_standard_bams_dmp - curated DMP standard BAM from NovaSeq

  • hiseq4000_curated_standard_bams_dmp - curated DMP standard BAM from HiSeq 4000

  • novaseq_curated_unfiltered_bams_dmp - curated DMP unfiltered BAM from NovaSeq

  • hiseq4000_curated_unfiltered_bams_dmp - curated DMP unfiltered BAM from HiSeq 4000

  • novaseq_unmatched_normal_plasma_duplex_bams_dmp - DMP unmatched normal plasma duplex BAM from NovaSeq

  • hiseq4000_unmatched_normal_plasma_duplex_bams_dmp - DMP unmatched normal plasma duplex BAM from HiSeq 4000

  • novaseq_unmatched_normal_plasma_standard_bams_dmp - DMP unmatched normal plasma standard BAM from NovaSeq

  • hiseq4000_unmatched_normal_plasma_standard_bams_dmp - DMP unmatched normal plasma standard BAM from HiSeq 4000

  • novaseq_curated_duplex_bams_dmp - curated DMP duplex BAMS from NovaSeq

  • regions_of_interest - Different interval files describing regions of interest for MSK-ACCESS assay

What is MSK-ACCESS?

It is a hybrid capture panel designed for Analysis of Circulating cfDNA to Evaluate Somatic Status using the Unique Molecular Index (UMIs) for high sensitivity. MSK-ACCESS is 13% as large, captures 47% of all mutations detected by MSK-IMPACT.

Brannon, A. R. et al. Enhanced specificity of clinical high-sensitivity tumor mutation profiling in cell- free DNA via paired normal sequencing using MSK-ACCESS. Nat Commun 12, 3770 (2021).

What is the panel design?

Selected exons of 129 genes for mutation detection

• OncoKB Level 1-4

• Hotspot sites

• High rates of mutations

• Protein kinase domains

• Tumor suppressor genes

Microsatellite regions

SNPs of zygosity & copy number of 12 genes

Common SNPs for genome-wide copy number

Introns for structural variants of 10 genes

Clonal hematopoiesis genes

What are the Design Implications?

• Matched cfDNA-WBC (”tumor-normal”) assay to detect somatic alterations

• Sensitivity for mutation calling depends on ‘duplex’ collapsed coverage

• Different sensitivities for different classes of alterations

  • Genotyping +++++

  • De novo mutations, indels ++++

  • MSI +++

  • Rearrangements +++

  • Copy number ++

  • Tumor mutation burden ○

SMILE -> information associated with a sample. This is a current work-in-progress written in Java/Neo4J that is supposed to be the one source of truth for metadata associated with samples. Currently, these responsibilities are managed by Beagle.

LIMS -> Laboratory information management system -- when sequencing is complete, metadata and file information on the sequences is first input into this system.

Beagle -> triggers and monitors workflows. There's also a part of it that tracks files and metadata associated with those files which may be splintered out in the future. A request typically comes from LIMS which begins the workflow process.

Hermes -> a nicer interface for Beagle (sample workflow tracking) and soon the MDB (make updates to metadata)

Ridgeback -> An HTTP API for Toil which Beagle is reliant on for interacting with workflows.

Toil -> A workflow engine that interfaces well with LSF.

Voyager -> The suite of applications built by the Voyager team, including Ridgeback, Beagle, and Hermes.

HPC-> Distributed high-performance computing that's managed in-house. Most data processing and files are housed here.

LSF -> IBM's Platform Computing (Load Sharing Facility) tool for scheduling workflows on HPC. It has a suite of commands specific to managing workloads.

Common workflow language -> a YAML-like language for defining workflows.

Nextflow -> Reactive workflow framework and a programming DSL that eases the writing of data-intensive computational pipelines.

DNA-SEQ -> Sequencing of the Deoxyribonucleic acid (DNA)

RNA-SEQ -> Sequencing of the Ribonucleic acid (RNA)

MSK-ACCESS -> Cell-Free DNA assay for patients with solid tumors

CMO-CH -> Assay for profiling clonal hematopoiesis mutations from blood

MSK-IMPACT -> Assay for profiling patients tissue DNA for solid tumors

MSK-IMPACT Heme -> Assay for profiling patients blood DNA for Heme malignancies

Facets -> Allele-specific copy number and clonal heterogeneity analysis tool for high-throughput DNA sequencing

WES -> Whole Exome Sequencing

WGS -> Whole Genome Sequencing

WTS -> Whole Transcriptome Sequencing

Illumina -> one of the companies that provides sequencing machine

Novaseq -> Type of sequencer from Illumina

Nextseq -> Type of sequencer from Illumina

Miseq -> Type of sequencer from Illumina

Hiseq -> Type of sequencer from Illumina

PacBio -> a company that sells instruments for long-read sequencing based on zero-mode waveguide

Oxford Nanopore -> a company that sells instruments for long-read sequencing based on Nanopore technology

To request time off, just fill in things at MSK TIME, and also please email your manager of the same.

This pages

Confluence knowledgebase across collabrators

CMO Project Managers

CCI works closely with CMO project managers to track the progress of cfDNA projects submitted to CMO by MSK researchers. CMO project managers support CCI by facilitating and coordinating project initiation, sample collection, and metadata collection.

People to connect

• Kirsten Fuller

• Casey Savin

• Jennifer Milbank

Integrated Genomics Operations (IGO)

CCI works with to make sure that the data generated for the cfDNA assays is of the highest standards.

People to connect

• Neeman Mohibullah

• Ruchi Patel

• Marisa Dunigan

• David Mcmanamon

CMO Informatics (CI)

CCI works closely with CMO Software Engineering (CSE) team within CI to support the development of Voyager & Hermes. The CSE/CAS team supports the development, integration, processing of CCI’s workflows implemented in Voyager.

People to connect

CMO Software Engineers (CSE)

• Christopher Allan Bolipata

• Sinisa Ivkovic

• Nikhil Kumar

CMO Analysis Systems (CAS)

• Timothy Song

• Stephen Kelly

• Suleman Vural

Berger Lab/Technology Innovation Lab/CMO Computational Science (CCS)

CCI collaborates with scientists in various CMO groups to improve the existing workflows and to analyze the data in a consistent manner.

People to connect

Berger Lab

• Kanika Arora

• Shalabh Suman

• Chaitanya Bandlamundi

Technology Innovation Lab

• Brian Loomis

• David Brown

CMO Computational Sciences (CCS)

• Mark Donoghue

• Yixiao Gong

Clinical Bioinformatics (ClinBx)

CCI works closely with the ClinBx group in the Molecular Diagnostics Service to deploy core workflows and maintain consistency among analyses performed on research and clinical cfDNA samples. We work together to improve the workflows in sync and to learn from one another about how to identify artifacts and interpret the data. CCI is also working with the ClinBx Software group to port an instance of the mPATH system used by the ClinBx team to sign-out cases for research.

People to know

Rose Brannon

Aijazuddin Syed

Ryan Ptashkin

Anoop Balakrishnan

High Performance Computing (HPC)

CCI works with HPC to request resources and support w.r.t JUNO cluster and virtual machines that enable various aspects of our goals.

Neeraj Girija

Lohit Valleru

Slack Channel

hpc-request@cbio.mskcc.org

To get reimbursed for your expenses, just fill in the expense report online.