Phenotype prediction from single-cell RNA-seq data using attention-based neural networks.

MOTIVATION: A patient's disease phenotype can be driven and determined by specific groups of cells whose marker genes are either unknown or can only be detected at late-stage using conventional bulk assays such as RNA-Seq technology. Recent advances in single-cell RNA sequencing (scRNA-seq) enable gene expression profiling in cell-level resolution, and therefore have the potential to identify those cells driving the disease phenotype even while the number of these cells is small. However, most existing methods rely heavily on accurate cell type detection, and the number of available annotated samples is usually too small for training deep learning predictive models. RESULTS: Here, we propose the method ScRAT for phenotype prediction using scRNA-seq data. To train ScRAT with a limited number of samples of different phenotypes, such as coronavirus disease (COVID) and non-COVID, ScRAT first applies a mixup module to increase the number of training samples. A multi-head attention mechanism is employed to learn the most informative cells for each phenotype without relying on a given cell type annotation. Using three public COVID datasets, we show that ScRAT outperforms other phenotype prediction methods. The performance edge of ScRAT over its competitors increases as the number of training samples decreases, indicating the efficacy of our sample mixup. Critical cell types detected based on high-attention cells also support novel findings in the original papers and the recent literature. This suggests that ScRAT overcomes the challenge of missing marker genes and limited sample number with great potential revealing novel molecular mechanisms and/or therapies. AVAILABILITY AND IMPLEMENTATION: The code of our proposed method ScRAT is published at https://github.com/yuzhenmao/ScRAT.

Optimized high-throughput screening of non-coding variants identified from genome-wide association studies.

The vast majority of disease-associated single nucleotide polymorphisms (SNP) identified from genome-wide association studies (GWAS) are localized in non-coding regions. A significant fraction of these variants impact transcription factors binding to enhancer elements and alter gene expression. To functionally interrogate the activity of such variants we developed snpSTARRseq, a high-throughput experimental method that can interrogate the functional impact of hundreds to thousands of non-coding variants on enhancer activity. snpSTARRseq dramatically improves signal-to-noise by utilizing a novel sequencing and bioinformatic approach that increases both insert size and the number of variants tested per loci. Using this strategy, we interrogated known prostate cancer (PCa) risk-associated loci and demonstrated that 35% of them harbor SNPs that significantly altered enhancer activity. Combining these results with chromosomal looping data we could identify interacting genes and provide a mechanism of action for 20 PCa GWAS risk regions. When benchmarked to orthogonal methods, snpSTARRseq showed a strong correlation with in vivo experimental allelic-imbalance studies whereas there was no correlation with predictive in silico approaches. Overall, snpSTARRseq provides an integrated experimental and computational framework to functionally test non-coding genetic variants.

Isolating Nuclei From Frozen Human Heart Tissue for Single-Nucleus RNA Sequencing.

Heart disease is the leading cause of global morbidity and mortality. This is in part because, despite an abundance of animal and in vitro models, it has been a challenge to date to study human heart tissue with sufficient depth and resolution to develop disease-modifying therapies for common cardiac conditions. Single-nucleus RNA sequencing (snRNA-seq) has emerged as a powerful tool capable of analyzing cellular function and signaling in health and disease, and has already contributed to significant advances in areas such as oncology and hematology. Employing snRNA-seq technology on flash-frozen human tissue has the potential to unlock novel disease mechanisms and pathways in any organ. Studying the human heart using snRNA-seq is a key priority for the field of cardiovascular sciences; however, progress to date has been slowed by numerous barriers. One key challenge is the fact that the human heart is very resistant to shearing and stress, making tissue dissociation and nuclear isolation difficult. Here, we describe a tissue dissociation method allowing the efficient and cost-effective isolation of high-quality nuclei from flash-frozen human heart tissue collected in surgical operating rooms. Our protocol addresses the challenge of nuclear isolation from human hearts, enables snRNA-seq of the human heart, and paves the way for an improved understanding of the human heart in health and disease. Ultimately, this will be key to uncovering signaling pathways and networks amenable to therapeutic intervention and the development of novel biomarkers and disease-modifying therapies. © 2022 Wiley Periodicals LLC. Basic Protocol: Human heart tissue dissociation and nuclear isolation for snRNA-seq.

Functional mapping of androgen receptor enhancer activity.

BACKGROUND: Androgen receptor (AR) is critical to the initiation, growth, and progression of prostate cancer. Once activated, the AR binds to cis-regulatory enhancer elements on DNA that drive gene expression. Yet, there are 10-100× more binding sites than differentially expressed genes. It is unclear how or if these excess binding sites impact gene transcription. RESULTS: To characterize the regulatory logic of AR-mediated transcription, we generated a locus-specific map of enhancer activity by functionally testing all common clinical AR binding sites with Self-Transcribing Active Regulatory Regions sequencing (STARRseq). Only 7% of AR binding sites displayed androgen-dependent enhancer activity. Instead, the vast majority of AR binding sites were either inactive or constitutively active enhancers. These annotations strongly correlated with enhancer-associated features of both in vitro cell lines and clinical prostate cancer samples. Evaluating the effect of each enhancer class on transcription, we found that AR-regulated enhancers frequently interact with promoters and form central chromosomal loops that are required for transcription. Somatic mutations of these critical AR-regulated enhancers often impact enhancer activity. CONCLUSIONS: Using a functional map of AR enhancer activity, we demonstrated that AR-regulated enhancers act as a regulatory hub that increases interactions with other AR binding sites and gene promoters.

Well-Differentiated Papillary Mesothelioma of the Peritoneum Is Genetically Distinct from Malignant Mesothelioma.

Well-differentiated papillary mesothelioma (WDPM) is an uncommon mesothelial proliferation that is most commonly encountered as an incidental finding in the peritoneal cavity. There is controversy in the literature about whether WDPM is a neoplasm or a reactive process and, if neoplastic, whether it is a variant or precursor of epithelial malignant mesothelioma or is a different entity. Using whole exome sequencing of five WDPMs of the peritoneum, we have identified distinct mutations in EHD1, ATM, FBXO10, SH2D2A, CDH5, MAGED1, and TP73 shared by WDPM cases but not reported in malignant mesotheliomas. Furthermore, we show that WDPM is strongly enriched with C > A transversion substitution mutations, a pattern that is also not found in malignant mesotheliomas. The WDPMs lacked the alterations involving BAP1, SETD2, NF2, CDKN2A/B, LASTS1/2, PBRM1, and SMARCC1 that are frequently found in malignant mesotheliomas. We conclude that WDPMs are neoplasms that are genetically distinct from malignant mesotheliomas and, based on observed mutations, do not appear to be precursors of malignant mesotheliomas.

A metabolomic strategy defines the regulation of lipid content and global metabolism by Δ9 desaturases in Caenorhabditis elegans.

BACKGROUND: Caenorhabditis elegans provides a genetically tractable model organism to investigate the network of genes involved in fat metabolism and how regulation is perturbed to produce the complex phenotype of obesity. C. elegans possess the full range of desaturases, including the Δ9 desaturases expressed by fat-5, fat-6 and fat-7. They regulate the biosynthesis of monounsaturated fatty acids, used for the synthesis of lipids including phospholipids, triglycerides and cholesteryl esters. RESULTS: Liquid chromatography mass spectrometry (LC-MS), gas chromatography mass spectrometry (GC-MS) and nuclear magnetic resonance (NMR) spectroscopy were used to define the metabolome of all the possible knock-outs for the Δ9 desaturases, including for the first time intact lipids. Despite the genes having similar enzymatic roles, excellent discrimination was achievable for all single and viable double mutants highlighting the distinctive roles of fat-6 and fat-7, both expressing steroyl-CoA desaturases. The metabolomic changes extend to aqueous metabolites demonstrating the influence Δ9 desaturases have on regulating global metabolism and highlighting how comprehensive metabolomics is more discriminatory than classically used dyes for fat staining. CONCLUSIONS: The propagation of metabolic changes across the network of metabolism demonstrates that modification of the Δ9 desaturases places C.elegans into a catabolic state compared with wildtype controls.

Human claspin is a ring-shaped DNA-binding protein with high affinity to branched DNA structures.

Claspin is an essential protein for the ATR-dependent activation of the DNA replication checkpoint response in Xenopus and human cells. Here we describe the purification and characterization of human Claspin. The protein has a ring-like structure and binds with high affinity to branched DNA molecules. These findings suggest that Claspin may be a component of the replication ensemble and plays a role in the replication checkpoint by directly associating with replication forks and with the various branched DNA structures likely to form at stalled replication forks because of DNA damage.

Molecular characterization of a full genome Turkish hepatitis C virus 1b isolate (HCV-TR1): a predominant viral form in Turkey.

Based on direct sequencing information from 5'UTR and NS5B regions, we identified subtype lb as a predominant hepatitis C virus genome in Turkey, which affected more than 91% of 79 patients studied. Next, the full genome sequence of a Turkish lb isolate was obtained by the cloning of polypeptide-encoding region into 7 overlapping fragments. Turkish 1b isolate, which was named HCV-TR1, comprises 9361 nucleotides, including 306 nucleotides of 5'UTR, a single long open reading frame of 9033 nucleotides, and 22 nucleotides of 3'UTR. When compared to HCV lb polypeptide sequences available at GenBank, the predicted polypeptide displayed a total of 36 amino acid substitutions, of which 16 was specific for HCV-TR1 isolate. Despite these changes, major structural and functional motifs of HCV proteins were maintained in HCV-TR1. In contrast, HCV-TR1 displayed amino acid substitutions in 6 out of 9 major cytotoxic T-cell epitopes. These data suggest that HCV-TR1 encodes functionally intact viral proteins, but it also encodes altered viral epitopes, which may affect host immune-response.