Orthrus: Towards Evolutionary and Functional RNA Foundation Models.

In the face of rapidly accumulating genomic data, our ability to accurately predict key mature RNA properties that underlie transcript function and regulation remains limited. Pre-trained genomic foundation models offer an avenue to adapt learned RNA representations to biological prediction tasks. However, existing genomic foundation models are trained using strategies borrowed from textual or visual domains that do not leverage biological domain knowledge.

Identification of genetic subtypes in follicular lymphoma.

Follicular lymphoma (FL) exhibits considerable variability in biological features and clinical trajectories across patients. To dissect the diversity of FL, we utilized a Bernoulli mixture model to identify genetic subtypes in 713 pre-treatment tumor tissue samples. Our analysis revealed the existence of five subtypes with unique genetic profiles that correlated with clinicopathological characteristics.

Cas13d-mediated isoform-specific RNA knockdown with a unified computational and experimental toolbox.

Alternative splicing is an essential mechanism for diversifying proteins, in which mature RNA isoforms produce proteins with potentially distinct functions. Two major challenges in characterizing the cellular function of isoforms are the lack of experimental methods to specifically and efficiently modulate isoform expression and computational tools for complex experimental design. To address these gaps, we developed and methodically tested a strategy which pairs the RNA-targeting CRISPR/Cas13d system with guide RNAs that span exon-exon junctions in the mature RNA.

In-depth characterization of intratumoral heterogeneity in refractory B-cell non-Hodgkin lymphoma through the lens of a Research Autopsy Program.

Intratumoral heterogeneity (ITH) provides the substrate for tumor evolution and treatment resistance, yet is remarkably understudied in lymphoma, due to the often limited amount of tissue that gets sampled during the routine diagnostic process, generally from a single nodal or extranodal site. Furthermore, the trajectory of how lymphoma, and especially non-Hodgkin lymphoma, spreads throughout the human body remains poorly understood. Here, we present a detailed characterization of ITH by applying whole-genome sequencing to spatially separated tumor samples harvested at the time of autopsy (n=24) and/or diagnosis (n=3) in three patients presenting with refractory B-cell non-Hodgkin lymphoma.

Pan-cancer analysis of non-coding transcripts reveals the prognostic onco-lncRNA HOXA10-AS in gliomas.

Long non-coding RNAs (lncRNAs) are increasingly recognized as functional units in cancer and powerful biomarkers; however, most remain uncharacterized. Here, we analyze 5,592 prognostic lncRNAs in 9,446 cancers of 30 types using machine learning. We identify 166 lncRNAs whose expression correlates with survival and improves the accuracy of common clinical variables, molecular features, and cancer subtypes.

Combined EZH2 Inhibition and IKAROS Degradation Leads to Enhanced Antitumor Activity in Diffuse Large B-cell Lymphoma.

Purpose: The efficacy of EZH2 inhibition has been modest in the initial clinical exploration of diffuse large B-cell lymphoma (DLBCL), yet EZH2 inhibitors are well tolerated. Herein, we aimed to uncover genetic and pharmacologic opportunities to enhance the clinical efficacy of EZH2 inhibitors in DLBCL.

Experimental Design: We conducted a genome-wide sensitizing CRISPR/Cas9 screen with tazemetostat, a catalytic inhibitor of EZH2.

A risk model for relapsed/refractory aggressive NHL integrating clinical risk factors and pretransplant Deauville score.

There are limited data regarding the combined value of the pretransplant Deauville score (DS) from a positron emission tomography scan and clinical risk factors in patients with relapsed/refractory aggressive non-Hodgkin lymphoma (NHL). We performed a retrospective analysis to assess the prognostic role of pretransplant DS in patients with relapsed/refractory aggressive NHL who underwent salvage chemotherapy and autologous stem cell transplantation (ASCT). We identified 174 eligible patients between January 2013 and March 2019.

Analyses of non-coding somatic drivers in 2,658 cancer whole genomes.

The discovery of drivers of cancer has traditionally focused on protein-coding genes. Here we present analyses of driver point mutations and structural variants in non-coding regions across 2,658 genomes from the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium of the International Cancer Genome Consortium (ICGC) and The Cancer Genome Atlas (TCGA). For point mutations, we developed a statistically rigorous strategy for combining significance levels from multiple methods of driver discovery that overcomes the limitations of individual methods.

Candidate Cancer Driver Mutations in Distal Regulatory Elements and Long-Range Chromatin Interaction Networks.

A comprehensive catalog of cancer driver mutations is essential for understanding tumorigenesis and developing therapies. Exome-sequencing studies have mapped many protein-coding drivers, yet few non-coding drivers are known because genome-wide discovery is challenging. We developed a driver discovery method, ActiveDriverWGS, and analyzed 120,788 cis-regulatory modules (CRMs) across 1,844 whole tumor genomes from the ICGC-TCGA PCAWG project.

ID1 Is Critical for Tumorigenesis and Regulates Chemoresistance in Glioblastoma.

Glioblastoma is the most common primary brain tumor in adults. While the introduction of temozolomide chemotherapy has increased long-term survivorship, treatment failure and rapid tumor recurrence remains universal. The transcriptional regulatory protein, inhibitor of DNA-binding-1 (ID1), is a key regulator of cell phenotype in cancer.

Intertumoral Heterogeneity within Medulloblastoma Subgroups.

While molecular subgrouping has revolutionized medulloblastoma classification, the extent of heterogeneity within subgroups is unknown. Similarity network fusion (SNF) applied to genome-wide DNA methylation and gene expression data across 763 primary samples identifies very homogeneous clusters of patients, supporting the presence of medulloblastoma subtypes. After integration of somatic copy-number alterations, and clinical features specific to each cluster, we identify 12 different subtypes of medulloblastoma.

Noncoding RNA for personalized prostate cancer treatment: utilizing the 'dark matters' of the genome.

Prostate cancer is the most commonly diagnosed cancer in men in western countries, with significant health impact. Clinically, it is complicated with the lack of biomarkers and effective treatments for aggressive disease, particularly castration-resistant prostate cancer. Although we have gained much insight into the biology of prostate cancer through studying protein-coding genes, they represent only a small fraction of our genome.