Demo Faculty

Chromatin immunoprecipitation sequencing (ChIP-seq) and the Assay for Transposase-Accessible Chromatin with high-throughput sequencing (ATAC-seq) have become essential technologies to effectively measure protein–DNA interactions and chromatin accessibility. However, there is a need for a scalable and reproducible pipeline that incorporates proper normalization between samples, correction of copy number variations, and integration of new downstream analysis tools. Here we present Containerized Bioinformatics workflow for Reproducible ChIP/ATAC-seq Analysis (CoBRA), a modularized computational workflow which quantifies ChIP-seq and ATAC-seq peak regions and performs unsupervised and supervised analyses. CoBRA provides a comprehensive state-of-the-art ChIP-seq and ATAC-seq analysis pipeline that can be used by scientists with limited computational experience. This enables …

Relapsed myeloid disease after allogeneic stem cell transplantation (HSCT) remains largely incurable. We previously demonstrated the potent activity of immune checkpoint blockade in this clinical setting with ipilimumab or nivolumab. To define the molecular and cellular pathways by which CTLA-4 blockade with ipilimumab can reinvigorate an effective graft-versus-leukemia (GVL) response, we integrated transcriptomic analysis of leukemic biopsies with immunophenotypic profiling of matched peripheral blood samples collected from patients treated with ipilimumab following HSCT on the Experimental Therapeutics Clinical Trials Network 9204 trial. Response to ipilimumab was associated with transcriptomic evidence of increased local CD8+ T-cell infiltration and activation. Systemically, ipilimumab decreased naïve and increased memory T-cell populations and increased expression of markers of T-cell …

We introduce the TRUST4 open-source algorithm for reconstruction of immune receptor repertoires in αβ/γδ T cells and B cells from RNA-sequencing (RNA-seq) data. Compared with competing methods, TRUST4 supports both FASTQ and BAM format and is faster and more sensitive in assembling longer—even full-length—receptor repertoires. TRUST4 can also call repertoire sequences from single-cell RNA-seq (scRNA-seq) data without V (D) J enrichment, and is compatible with both SMART-seq and 5′ 10x Genomics platforms.

Deep learning applied to genomics can learn patterns in biological sequences, but designing such models requires expertise and effort. Recent work demonstrates the efficiency of a neural network architecture search algorithm in optimizing genomic models.

Deep learning has been powerful in learning complex functions from data and has been applied in computer vision, natural language processing and biology. If we view the human genome as a book with three billion letters of nucleotides represented by A, C, G and T, genes and gene-controlling sequences are encoded in the book and variations in the genome can link to disease conditions. Neural network models that extract patterns from the sequences can help predict functional genomic elements and interpret genetic variations. However, the current deep learning models for genomics usually involve expert-designed neural network structures and require extensive tuning, making such models unapproachable for most other scientists. In a recent publication in Nature Machine Intelligence, Zhang and colleagues 1 present a framework called Automated Modelling for Biological Evidence-based Research …

Background Large-scale genome-wide association studies (GWAS) have implicated thousands of germline genetic variants in modulating individuals’ risk to various diseases, including cancer. At least 25 risk loci have been identified for low-grade gliomas (LGGs), but their molecular functions remain largely unknown. Methods We hypothesized that GWAS loci contain causal single nucleotide polymorphisms (SNPs) that reside in accessible open chromatin regions and modulate the expression of target genes by perturbing the binding affinity of transcription factors (TFs). We performed an integrative analysis of genomic and epigenomic data from The Cancer Genome Atlas and other public repositories to identify candidate causal SNPs within linkage disequilibrium blocks of LGG GWAS loci. We assessed their potential regulatory role via in silico TF binding sequence …

MotivationThe chromatin profile measured by ATAC-seq, ChlP-seq, or DNase-seq experiments can identify genomic regions critical in regulating gene expression and provide insights on biological processes such as diseases and development. However, quality control and processing chromatin profiling data involve many steps, and different bioinformatics tools are used at each step. It can be challenging to manage the analysis.ResultsWe developed a Snakemake pipeline called CHIPS (CHromatin enrichment Processor) to streamline the processing of ChIP-seq, ATAC-seq, and DNase-seq data. The pipeline supports single- and paired-end data and is flexible to start with FASTQ or BAM files. It includes basic steps such as read trimming, mapping, and peak calling. In addition, it calculates quality control metrics such as contamination profiles, PCR bottleneck coefficient, the fraction of reads in peaks, percentage of peaks overlapping with the union of public DNaseI hypersensitivity sites, and conservation profile of the peaks. For downstream analysis, it carries out peak annotations, motif finding, and regulatory potential calculation for all genes. The pipeline ensures that the processing is robust and reproducible.AvailabilityCHIPS is available at https://bitbucket.org/plumbers/cidc_chips/src/master/Contactmtang@ds.dfci.harvard.edu: henry_long@dfci.harvard.edu

Creating immersive 3D stereoscopic, autostereoscopic, and lightfield experiences are becoming the center point of optical design of future head mounted displays and lightfield displays. However, despite the advancement in 3D and light field displays, there is no consensus on what are the necessary quantized depth levels for such emerging displays at stereoscopic or monocular modalities. Here we start from psychophysical theories and work toward defining and prioritizing quantized levels of depth that would saturate the human depth perception. We propose a general optimization framework, which locates the depth levels in a globally optimal way for band limited displays. While the original problem is computationally intractable, we manage to find a tractable reformulation as maximally covering a region of interest with a selection of hypographs corresponding to the monocular depth of field profiles. The results …

The focusing response of the human eye—accommodation—exhibits errors known as lags and leads. Lags occur when the stimulus is near and the eye appears to focus farther than the stimulus. Leads occur with far stimuli where the eye appears to focus nearer than the stimulus. We used objective and subjective measures simultaneously to determine where the eye is best focused. The objective measures were made with a wavefront sensor and an autorefractor, both of which analyze light reflected from the retina. These measures exhibited typical accommodative errors, mostly lags. The subjective measure was visual acuity, which of course depends not only on the eye’s optics but also on photoreception and neural processing of the retinal image. The subjective measure revealed much smaller errors. Acuity was maximized at or very close to the distance of the accommodative stimulus. Thus, accommodation is accurate in terms of maximizing visual performance.

T cells are critical effectors of cancer immunotherapies, but little is known about their gene expression programs in diffuse gliomas. Here, we leverage single-cell RNA sequencing (RNA-seq) to chart the gene expression and clonal landscape of tumor-infiltrating T cells across 31 patients with isocitrate dehydrogenase (IDH) wild-type glioblastoma and IDH mutant glioma. We identify potential effectors of anti-tumor immunity in subsets of T cells that co-express cytotoxic programs and several natural killer (NK) cell genes. Analysis of clonally expanded tumor-infiltrating T cells further identifies the NK gene KLRB1 (encoding CD161) as a candidate inhibitory receptor. Accordingly, genetic inactivation of KLRB1 or antibody-mediated CD161 blockade enhances T cell-mediated killing of glioma cells in vitro and their anti-tumor function in vivo. KLRB1 and its associated transcriptional program are also expressed by …

The ubiquitin-proteasome system (UPS) is the primary route for selective protein degradation in human cells. The UPS is an attractive target for novel cancer therapies, but the precise UPS genes and substrates important for cancer growth are incompletely understood. Leveraging multi-omics data across more than 9,000 human tumors and 33 cancer types, we found that over 19% of all cancer driver genes affect UPS function. We implicate transcription factors as important substrates and show that c-Myc stability is modulated by CUL3. Moreover, we developed a deep learning model (deepDegron) to identify mutations that result in degron loss and experimentally validated the prediction that gain-of-function truncating mutations in GATA3 and PPM1D result in increased protein stability. Last, we identified UPS driver genes associated with prognosis and the tumor microenvironment. This study demonstrates the …

Cancer immunotherapy shows limited efficacy against many solid tumors that originate from epithelial tissues, including triple-negative breast cancer (TNBC). We identify the SOX4 transcription factor as an important resistance mechanism to T cell-mediated cytotoxicity for TNBC cells. Mechanistic studies demonstrate that inactivation of SOX4 in tumor cells increases the expression of genes in a number of innate and adaptive immune pathways important for protective tumor immunity. Expression of SOX4 is regulated by the integrin αvβ6 receptor on the surface of tumor cells, which activates TGFβ from a latent precursor. An integrin αvβ6/8-blocking monoclonal antibody (mAb) inhibits SOX4 expression and sensitizes TNBC cells to cytotoxic T cells. This integrin mAb induces a substantial survival benefit in highly metastatic murine TNBC models poorly responsive to PD-1 blockade. Targeting of the integrin αvβ6-TGFβ …

Immune checkpoint blockade (ICB) therapy revolutionized cancer treatment, but many patients with impaired MHC-I expression remain refractory. Here, we combined FACS-based genome-wide CRISPR screens with a data-mining approach to identify drugs that can upregulate MHC-I without inducing PD-L1. CRISPR screening identified TRAF3, a suppressor of the NFκB pathway, as a negative regulator of MHC-I but not PD-L1. The Traf3-knockout gene expression signature is associated with better survival in ICB-naïve patients with cancer and better ICB response. We then screened for drugs with similar transcriptional effects as this signature and identified Second Mitochondria-derived Activator of Caspase (SMAC) mimetics. We experimentally validated that the SMAC mimetic birinapant upregulates MHC-I, sensitizes cancer cells to T cell–dependent killing, and adds to ICB efficacy. Our findings provide …

Understanding the impact of regulatory variants on complex phenotypes is a significant challenge because the genes and pathways that are targeted by such variants are typically unknown. Furthermore, a regulatory variant might influence a particular gene9s expression in a cell type or tissue-specific manner. Cell-type specific long-range regulatory interactions that occur between a distal regulatory sequence and a gene offers a powerful framework for understanding the impact of regulatory variants on complex phenotypes. However, high-resolution maps of such long-range interactions are available only for a handful of model cell lines. To address this challenge, we have developed L-HiC-Reg, a Random Forests based regression method to predict high-resolution contact counts in new cell lines, and a network-based framework to identify candidate cell line-specific gene networks targeted by a set of variants from a Genome-wide association study (GWAS). We applied our approach to predict interactions in 55 Roadmap Epigenome Consortium cell lines, which we used to interpret regulatory SNPs in the NHGRI GWAS catalogue. Using our approach, we performed an in-depth characterization of fifteen different phenotypes including Schizophrenia, Coronary Artery Disease (CAD) and Crohn9s disease. In CAD, we found differentially wired subnetworks consisting of known as well as novel gene targets of regulatory SNPs. Taken together, our compendium of interactions and associated network-based analysis pipeline offers a powerful resource to leverage long-range regulatory interactions to examine the context-specific impact of regulatory …

A number of cancer drugs activate innate immune pathways in tumor cells but unfortunately also compromise anti-tumor immune function. We discovered that inhibition of Carm1, an epigenetic enzyme and co-transcriptional activator, elicited beneficial anti-tumor activity in both cytotoxic T cells and tumor cells. In T cells, Carm1 inactivation substantially enhanced their anti-tumor function and preserved memory-like populations required for sustained anti-tumor immunity. In tumor cells, Carm1 inactivation induced a potent type 1 interferon response that sensitized resistant tumors to cytotoxic T cells. Substantially increased numbers of dendritic cells, CD8 T cells and NK cells were present in Carm1-deficient tumors, and infiltrating CD8 T cells expressed low levels of exhaustion markers. Targeting of Carm1 with a small molecule elicited potent anti-tumor immunity and sensitized resistant tumors to checkpoint blockade …

Understanding the impact of regulatory variants on complex phenotypes is a significant challenge because the genes and pathways that are targeted by such variants are typically unknown. Furthermore, a regulatory variant might influence a particular gene9s expression in a cell type or tissue-specific manner. Cell-type specific long-range regulatory interactions that occur between a distal regulatory sequence and a gene offers a powerful framework for understanding the impact of regulatory variants on complex phenotypes. However, high-resolution maps of such long-range interactions are available only for a handful of model cell lines. To address this challenge, we have developed L-HiC-Reg, a Random Forests based regression method to predict high-resolution contact counts in new cell lines, and a network-based framework to identify candidate cell line-specific gene networks targeted by a set of variants from a Genome-wide association study (GWAS). We applied our approach to predict interactions in 55 Roadmap Epigenome Consortium cell lines, which we used to interpret regulatory SNPs in the NHGRI GWAS catalogue. Using our approach, we performed an in-depth characterization of fifteen different phenotypes including Schizophrenia, Coronary Artery Disease (CAD) and Crohn9s disease. In CAD, we found differentially wired subnetworks consisting of known as well as novel gene targets of regulatory SNPs. Taken together, our compendium of interactions and associated network-based analysis pipeline offers a powerful resource to leverage long-range regulatory interactions to examine the context-specific impact of regulatory …

A number of cancer drugs activate innate immune pathways in tumor cells but unfortunately also compromise antitumor immune function. We discovered that inhibition of CARM1, an epigenetic enzyme and cotranscriptional activator, elicited beneficial antitumor activity in both cytotoxic T cells and tumor cells. In T cells, Carm1 inactivation substantially enhanced their antitumor function and preserved memory-like populations required for sustained antitumor immunity. In tumor cells, Carm1 inactivation induced a potent type 1 interferon response that sensitized resistant tumors to cytotoxic T cells. Substantially increased numbers of dendritic cells, CD8 T cells, and natural killer cells were present in Carm1-deficient tumors, and infiltrating CD8 T cells expressed low levels of exhaustion markers. Targeting of CARM1 with a small molecule elicited potent antitumor immunity and sensitized resistant tumors …

Blood malignancies provide unique opportunities for longitudinal tracking of disease evolution following therapeutic bottlenecks and for the monitoring of changes in anti-tumor immunity. The expanding development of multi-modal single-cell sequencing technologies affords newer platforms to elucidate the mechanisms underlying these processes at unprecedented resolution. Furthermore, the identification of molecular events that can serve as in-vivo barcodes now facilitate the tracking of the trajectories of malignant and of immune cell populations over time within primary human samples, as these permit unambiguous identification of the clonal lineage of cell populations within heterogeneous phenotypes. Here, we provide an overview of the potential for chromosomal copy number changes, somatic nuclear and mitochondrial DNA mutations, single nucleotide polymorphisms, and T and B cell receptor sequences …

Targeted protein degradation (TPD) has rapidly emerged as a therapeutic modality to eliminate previously undruggable proteins by repurposing the cell9s endogenous protein degradation machinery. However, the susceptibility of proteins for targeting by TPD approaches, termed "degradability", is largely unknown. Recent systematic studies to map the degradable kinome have shown differences in degradation between kinases with similar drug-target engagement, suggesting yet unknown factors influencing degradability. We therefore developed a machine learning model, MAPD (Model-based Analysis of Protein Degradability), to predict degradability from protein features that encompass post-translational modifications, protein stability, protein expression and protein-protein interactions. MAPD shows accurate performance in predicting kinases that are degradable by TPD compounds (auPRC=0.759) and is likely generalizable to independent non-kinase proteins. We found five features with statistical significance to achieve optimal prediction, with ubiquitination potential being the most predictive. By structural modeling, we found that E2-accessible ubiquitination sites, but not lysine residues in general, are particularly associated with kinase degradability. Finally, we extended MAPD predictions to the entire proteome to find 964 disease-causing proteins, including 278 cancer genes, that may be tractable to TPD drug development.

c-MYC (MYC) is a major driver of prostate cancer tumorigenesis and progression. Although MYC is overexpressed in both early and metastatic disease and associated with poor survival, its impact on prostate transcriptional reprogramming remains elusive. We demonstrate that MYC overexpression significantly diminishes the androgen receptor (AR) transcriptional program (the set of genes directly targeted by the AR protein) in luminal prostate cells without altering AR expression. Importantly, analyses of clinical specimens revealed that concurrent low AR and high MYC transcriptional programs accelerate prostate cancer progression toward a metastatic, castration-resistant disease. Data integration of single-cell transcriptomics together with ChIP-seq revealed an increased RNA polymerase II (Pol II) promoter-proximal pausing at AR-dependent genes following MYC overexpression without an accompanying deactivation of AR-bound enhancers. Altogether, our findings suggest that MYC overexpression antagonizes the canonical AR transcriptional program and contributes to prostate tumor initiation and progression by disrupting transcriptional pause release at AR-regulated genes.