Breast adenoid cystic carcinoma (AdCC), a rare type of triple-negative breast cancer, has been shown to be driven by MYB pathway activation, most often underpinned by the MYB-NFIB fusion gene. Alternative genetic mechanisms, such as MYBL1 rearrangements, have been reported in MYB-NFIB-negative salivary gland AdCCs. Here we report on the molecular characterization by massively parallel sequencing of four breast AdCCs lacking the MYB-NFIB fusion gene. In two cases, we identified MYBL1 rearrangements (MYBL1-ACTN1 and MYBL1-NFIB), which were associated with MYBL1 overexpression. A third AdCC harboured a high-level MYB amplification, which resulted in MYB overexpression at the mRNA and protein levels. RNA-sequencing and whole-genome sequencing revealed no definite alternative driver in the fourth AdCC studied, despite high levels of MYB expression and the activation of pathways similar to those activated in MYB-NFIB-positive AdCCs. In this case, a deletion encompassing the last intron and part of exon 15 of MYB, including the binding site of ERG-1, a transcription factor that may downregulate MYB, and the exon 15 splice site, was detected. In conclusion, we demonstrate that MYBL1 rearrangements and MYB amplification probably constitute alternative genetic drivers of breast AdCCs, functioning through MYBL1 or MYB overexpression. These observations emphasize that breast AdCCs probably constitute a convergent phenotype, whereby activation of MYB and MYBL1 and their downstream targets can be driven by the MYB-NFIB fusion gene, MYBL1 rearrangements, MYB amplification, or other yet to be identified mechanisms. Copyright © 2017 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

We report a novel computational method, RegNetDriver, to identify tumorigenic drivers using the combined effects of coding and non-coding single nucleotide variants, structural variants, and DNA methylation changes in the DNase I hypersensitivity based regulatory network. Integration of multi-omics data from 521 prostate tumor samples indicated a stronger regulatory impact of structural variants, as they affect more transcription factor hubs in the tissue-specific network. Moreover, crosstalk between transcription factor hub expression modulated by structural variants and methylation levels likely leads to the differential expression of target genes. We report known prostate tumor regulatory drivers and nominate novel transcription factors (ERF, CREB3L1, and POU2F2), which are supported by functional validation.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-017-1266-3) contains supplementary material, which is available to authorized users.

Most cancer types lack effective targeted therapeutic options and in cancers where first-line targeted therapies are available, treatment resistance is a huge challenge. Recent technological advances enable the use of ATAC-seq and RNA-seq on patient biopsies in a high-throughput manner. Here we present a computational approach that leverages these datasets to identify novel drug targets based on tumor lineage. We constructed patient-specific gene regulatory networks for 371 patients of 22 cancer types using machine learning approaches trained using three-dimensional genomic data for enhancer to promoter contacts. Next, we identify the key transcription factors (TFs) in these networks, which are used to identify therapeutic vulnerabilities either by direct targeting of TFs or proteins that they co-operate with. We validate four novel candidates identified for neuroendocrine, liver and renal cancers, which have a dismal prognosis with current therapeutic options. We present a novel approach to use the increasing amounts of functional genomics data from patient biospecimens for identification of novel drug targets.

Regulatory networks containing enhancer to gene edges define cellular state and their rewiring is a hallmark of cancer. While efforts, such as ENCODE, have revealed these networks for reference tissues and cell-lines by integrating multi-omics data, the same methods cannot be applied for large patient cohorts due to the constraints on generating ChIP-seq and three-dimensional data from limited material in patient biopsies. We trained a supervised machine learning model using genomic 3D signatures of physical enhancer-gene connections that can predict accurate connections using data from ATAC-seq and RNA-seq assays only, which can be easily generated from patient biopsies. Our method overcomes the major limitations of correlation-based approaches that cannot distinguish between distinct target genes of given enhancers in different samples, which is a hallmark of network rewiring in cancer. Our model achieved an AUROC (area under receiver operating characteristic curve) of 0.91 and, importantly, can distinguish between active regulatory elements with connections to target genes and poised elements with no connections to target genes. Our predicted regulatory elements are validated by multi-omics data, including histone modification marks from ENCODE, with an average specificity of 0.92. Application of our model on chromatin accessibility and transcriptomic data from 400 cancer patients across 22 cancer types revealed novel cancer-type and subtype-specific enhancer-gene connections for known cancer genes. In one example, we identified two enhancers that regulate the expression of ESR1 in only ER+ breast cancer (BRCA) samples but not in ER-samples. These enhancers are predicted to contribute to the high expression of ESR1 in 93% of ER+ BRCA samples. Functional validation using CRISPRi confirms that inhibition of these enhancers decreases the expression of ESR1 in ER+ samples.

Regulatory networks containing enhancer to gene edges define cellular state and their rewiring is a hallmark of cancer. While efforts, such as ENCODE, have revealed these networks for reference tissues and cell-lines by integrating multi-omics data, the same methods cannot be applied for large patient cohorts due to the constraints on generating ChIP-seq and three-dimensional data from limited material in patient biopsies. Moreover, many cancer types lack effective targeted therapeutic options and in cancers where first-line targeted therapies are available, treatment resistance is a huge challenge, owing to both genetic and epigenetic alterations. Recent technological advances have enabled the use of ATAC-seq and RNA-seq on patient biopsies in a high-throughput manner. To tackle these problems, we trained a supervised machine learning model using genomic 3D signatures of physical enhancer-gene connections that can predict accurate connections using data from ATAC-seq and RNA-seq assays only. Using these data, we achieved an AUROC (area under receiver operating characteristic curve) of 0.91 for the identification of true regulatory element-gene connections and, importantly, can distinguish between active regulatory elements with connections to target genes and poised elements with no connections to target genes. Our predicted regulatory elements are validated by multi-omics data, including histone modification marks from ENCODE, with an average specificity of 0.92. Our model, applied on chromatin accessibility and transcriptomic data from 400 cancer patients across 22 cancer types revealed novel cancer-type and subtype-specific enhancer-gene connections for known cancer genes with experimental support for accurate prediction of subtype-specific enhancer target genes using CRISPRi in MCF7, T47D and MDA-MB-231 cell lines which represent hormone-receptor (HR) positive and HR- subtypes of breast cancer. We leverage these predictions to construct patient-specific gene regulatory networks, identify the key transcription factors (TFs) in these networks, clusters of patients with similar networks across cancer sites of origin and subsequently identify therapeutic vulnerabilities either by direct targeting of TFs or proteins that they co-operate with. We identify commonly used therapeutic agents for specific cancer types such as ESR1-targeting agents in ER+ breast cancer, KRAS and EGFR inhibitors in Lung and Colon cancers in addition to multiple novel therapeutic targets. We validated four novel candidates identified for neuroendocrine, liver and renal cancers, which have a dismal prognosis with current therapeutic options. Here we present a computational approach that combines multi-omics machine learning and network analysis then leverages these datasets to identify novel drug targets based on tumor lineage. Citation Format: Andre Neil Forbes, Duo Xu, Sandra Cohen, Ann Palladino, Priya Pancholi, Ekta Khurana. Discovery of novel therapeutic targets using 3D chromatin conformation and patient-specific gene regulatory networks [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2034.