Delineating yeast cleavage and polyadenylation signals using deep learning.

3'-end cleavage and polyadenylation is an essential process for eukaryotic mRNA maturation. In yeast species, the polyadenylation signals that recruit the processing machinery are degenerate and remain poorly characterized compared with the well-defined regulatory elements in mammals. Here we address this issue by developing deep learning models to deconvolute degenerate cis-regulatory elements and quantify their positional importance in mediating yeast poly(A) site formation, cleavage heterogeneity, and strength. In S. cerevisiae, cleavage heterogeneity is promoted by the depletion of U-rich elements around poly(A) sites as well as multiple occurrences of upstream UA-rich elements. Sites with high cleavage heterogeneity show overall lower strength. The site strength and tandem site distances modulate alternative polyadenylation (APA) under the diauxic stress. Finally, we develop a deep learning model to reveal the distinct motif configuration of S. pombe poly(A) sites, which show more precise cleavage than S. cerevisiae Altogether, our deep learning models provide unprecedented insights into poly(A) site formation of yeast species, and our results highlight divergent poly(A) signals across distantly related species.

Genome-wide mapping of 5-hydroxymethylcytosines in circulating cell-free DNA as a non-invasive approach for early detection of hepatocellular carcinoma.

OBJECTIVE: The lack of highly sensitive and specific diagnostic biomarkers is a major contributor to the poor outcomes of patients with hepatocellular carcinoma (HCC). We sought to develop a non-invasive diagnostic approach using circulating cell-free DNA (cfDNA) for the early detection of HCC. DESIGN: Applying the 5hmC-Seal technique, we obtained genome-wide 5-hydroxymethylcytosines (5hmC) in cfDNA samples from 2554 Chinese subjects: 1204 patients with HCC, 392 patients with chronic hepatitis B virus infection (CHB) or liver cirrhosis (LC) and 958 healthy individuals and patients with benign liver lesions. A diagnostic model for early HCC was developed through case-control analyses using the elastic net regularisation for feature selection. RESULTS: The 5hmC-Seal data from patients with HCC showed a genome-wide distribution enriched with liver-derived enhancer marks. We developed a 32-gene diagnostic model that accurately distinguished early HCC (stage 0/A) based on the Barcelona Clinic Liver Cancer staging system from non-HCC (validation set: area under curve (AUC)=88.4%; (95% CI 85.8% to 91.1%)), showing superior performance over α-fetoprotein (AFP). Besides detecting patients with early stage or small tumours (eg, ≤2.0 cm) from non-HCC, the 5hmC model showed high capacity for distinguishing early HCC from high risk subjects with CHB or LC history (validation set: AUC=84.6%; (95% CI 80.6% to 88.7%)), also significantly outperforming AFP. Furthermore, the 5hmC diagnostic model appeared to be independent from potential confounders (eg, smoking/alcohol intake history). CONCLUSION: We have developed and validated a non-invasive approach with clinical application potential for the early detection of HCC that are still surgically resectable in high risk individuals.

Delineating yeast cleavage and polyadenylation signals using deep learning.

3'-end cleavage and polyadenylation is an essential process for eukaryotic mRNA maturation. In yeast species, the polyadenylation signals that recruit the processing machinery are degenerate and remain poorly characterized compared to well-defined regulatory elements in mammals. Especially, recent deep sequencing experiments showed extensive cleavage heterogeneity for some mRNAs in Saccharomyces cerevisiae and uncovered the polyA motif differences between S. cerevisiae vs. Schizosaccharomyces pombe . The findings raised the fundamental question of how polyadenylation signals are formed in yeast. Here we addressed this question by developing deep learning models to deconvolute degenerate cis -regulatory elements and quantify their positional importance in mediating yeast polyA site formation, cleavage heterogeneity, and strength. In S. cerevisiae , cleavage heterogeneity is promoted by the depletion of U-rich elements around polyA sites as well as multiple occurrences of upstream UA-rich elements. Sites with high cleavage heterogeneity show overall lower strength. The site strength and tandem site distances modulate alternative polyadenylation (APA) under the diauxic stress. Finally, we developed a deep learning model to reveal the distinct motif configuration of S. pombe polyA sites which show more precise cleavage than S. cerevisiae . Altogether, our deep learning models provide unprecedented insights into polyA site formation across yeast species.

Deep learning of human polyadenylation sites at nucleotide resolution reveals molecular determinants of site usage and relevance in disease.

The genomic distribution of cleavage and polyadenylation (polyA) sites should be co-evolutionally optimized with the local gene structure. Otherwise, spurious polyadenylation can cause premature transcription termination and generate aberrant proteins. To obtain mechanistic insights into polyA site optimization across the human genome, we develop deep/machine learning models to identify genome-wide putative polyA sites at unprecedented nucleotide-level resolution and calculate their strength and usage in the genomic context. Our models quantitatively measure position-specific motif importance and their crosstalk in polyA site formation and cleavage heterogeneity. The intronic site expression is governed by the surrounding splicing landscape. The usage of alternative polyA sites in terminal exons is modulated by their relative locations and distance to downstream genes. Finally, we apply our models to reveal thousands of disease- and trait-associated genetic variants altering polyadenylation activity. Altogether, our models represent a valuable resource to dissect molecular mechanisms mediating genome-wide polyA site expression and characterize their functional roles in human diseases.

Phenotyping Multiple Organ Dysfunction Syndrome Using Temporal Trends in Critically Ill Children.

Multiple organ dysfunction syndrome (MODS) is one of the most common causes of death in critically ill children. However, despite decades of clinical trials, there are no comprehensive approaches to the management of MODS or effective targeted therapies that have consistently improved outcomes. Better understanding the heterogeneity of MODS and characterizing subgroups of MODS patients could improve our understanding of the syndrome and help us develop new management strategies. We analyzed a cohort of 5,297 children with MODS from two children's hospitals and used subgraph-augmented non-negative matrix factorization (SANMF) to identify unique temporal patterns in organ dysfunction across four novel subgroups. We demonstrate that these subgroups are composed of patients with distinct clinical characteristics and are independently predictive of clinical outcomes. Our work suggests that these subgroups represent four relevant phenotypes of pediatric MODS that could be used to identify novel management strategies.

Towards precision medicine: advances in 5-hydroxymethylcytosine cancer biomarker discovery in liquid biopsy.

Robust and clinically convenient biomarkers for cancer diagnosis, early detection, and prognosis have great potential to improve patient survival and are the key to precision medicine. The advent of next-generation sequencing technologies enables a more sensitive and comprehensive profiling of genetic and epigenetic information in tumor-derived materials. Researchers are now able to monitor the dynamics of tumorigenesis in new dimensions, such as using circulating cell-free DNA (cfDNA) and tumor DNA (ctDNA). Mutation-based assays in liquid biopsy cannot always provide consistent results across studies due partly to intra- and inter-tumoral heterogeneity as well as technical limitations. In contrast, epigenetic analysis of patient-derived cfDNA is a promising alternative, especially for early detection and disease surveillance, because epigenetic modifications are tissue-specific and reflect the dynamic process of cancer progression. Therefore, cfDNA-based epigenetic assays are emerging to be a highly sensitive, minimally invasive tool for cancer diagnosis and prognosis with great potential in future precise care of cancer patients. The major obstacle for applying epigenetic analysis of cfDNA, however, has been the lack of enabling techniques with high sensitivity and technical robustness. In this review, we summarized the advances in epigenome-wide profiling of 5-hydroxymethylcytosine (5hmC) in cfDNA, focusing on the detection approaches and potential role as biomarkers in different cancer types.