Computational biology and bioinformatics

Mitochondrial fatty acid synthesis (mtFAS) generates the precursor for lipoic acid synthesis, but the role of longer fatty acid products has remained unclear. Here, the authors generated an engineered variant of human 2E-enoyl-ACP reductase (MECR) of mtFAS to study the role of long chain fatty acids.

Alternative polyadenylation (APA) contributes to the post-transcriptional regulation of most human genes, yet the effects of APA are largely overlooked by conventional transcriptome-wide association studies (TWAS). Here, the authors conduct an APA-TWAS for 11 brain disorders, identifying hundreds of APA-linked disease susceptibility genes.

Cell type deconvolution in tumor spatial transcriptomics (ST) data remains challenging. Here, the authors develop Spatial Cellular Estimator for Tumors (SpaCET) to infer cell types and intercellular interactions from ST data in cancer across different platforms, with improved performance over similar methods.

The compound Rocaglamide A (RocA) is known for repressing translation initiation. Here the authors identify a dual mode of action for RocA in blocking translation initiation and elongation via eIF4A using previous datasets and new analyses.

Targeted proteomics enables robust hypothesis-driven research. Here, Yu et al. present a multiplexed approach for targeted pathway proteomics and apply it to quantify protein families across 480 fully genotyped Diversity Outbred mice, revealing impacts of genetic variation on protein expression and lipid metabolism.

Few genetic loci have been associated with tuberculosis infection, possibly because of the influence of genetic variation in the pathogen. Here, the authors integrate human and Mycobacterium tuberculosis genetics to find genome-genome interactions associated with infection.

Metagenomic Hi-C enables genome retrieval in microbial samples. Here, the authors develop an integrative method to recover complete viral genomes and detect virus-host pairs using metagenomic Hi-C data.

High-affinity antibodies are often identified through directed evolution but deep leaning methods hold great promise. Here the authors report RESP, a pipeline for efficient identification of high affinity antibodies, and apply this to the PD-L1 antibody Atezolizumab.

Understanding the heterogeneity of growth, response to therapy and progression dynamics in metastatic colorectal cancer (mCRC) remains critical. Here, the authors analyse lesion-specific response heterogeneity in 4,308 mCRC patients and find that organ-level progression sequence is associated with long-term survival.

Single-cell genomics has expanded to measure diverse molecular modalities within the same cell. Here the authors provide a computational framework called scTriangulate to integrate cluster annotations from diverse independent sources, algorithms, and modalities to define statistically stable populations.

A major challenge in analyzing scRNA-seq data arises from challenges related to dimensionality and the prevalence of dropout events. Here the authors develop a deep graph learning method called scMGCA based on a graph-embedding autoencoder that simultaneously learns cell-cell topology representation and cluster assignments, outperforming other state-of-the-art models across multiple platforms.

Many stem cells exhibit cell division coupled to differentiation, though the changes occurring between consecutive cell divisions have been difficult to study. Here they use synchronized hPSC culture to show that production of transcription factors and epigenetic changes are linked with cell division timing.

Trial data have shown that post-discharge malaria chemoprevention (PDMC) reduces the risk of readmission and death in children previously hospitalised with severe malarial anaemia. Here, the authors use mathematical modelling to estimate the potential epidemiological impacts of PDMC in malaria-endemic countries in Africa.

Many methods for single cell data integration have been developed, though mosaic integration remains challenging. Here the authors present scMoMaT, a mosaic integration method for single cell multi-modality data from multiple batches, that jointly learns cell representations and marker features across modalities for different cell clusters, to interpret the cell clusters from different modalities.

Conventional transcriptome-wide association study (TWAS) approaches predict genetically regulated gene expression at the tissue level. Here, the authors develop a framework for cell-type-aware TWAS that predicts cell-type level expression from genotype data and identifies disease-associated genes with cell-type-specific effects.

Richter syndrome (RS) is the transformation of chronic lymphocytic leukaemia (CLL) into aggressive lymphoma, in most cases diffuse large B-cell lymphoma (DLBCL). Here, the authors characterize the DNA methylation and transcriptomic profiles of RS samples, find a clonally-related CLL epigenetic imprint, and develop classifiers for “RS-type” de novo DLBCLs.

Comparing experimental mass spectra to reference spectra can enable natural product identification, but these spectral libraries are often incomplete and not universally applicable. Here, the authors present SNAP-MS, a tool that allows assigning compound families without experimental or calculated reference spectra.

Methods that perform data integration are needed to analyse spatial transcriptomics data from multiple tissue slides. Here, the authors present PRECAST, an efficient data integration method for multiple spatial transcriptomics datasets with complex batch or biological effects between slides.

Long-read sequencing is promising for the detection of structural variants (SVs), which requires algorithms with high sensitivity and precision. Here, the authors develop DeBreak, an algorithm for comprehensive and accurate SV detection in long-read sequencing data across different platforms, which outperforms other SV callers.

A key player in the formation of endoplasmic reticulum sheets is CLIMP-63, but mechanistic details remained elusive. Here authors combined cellular experiments and mathematical modelling to show that S-acylation of CLIMP-63 regulates its function by mediating its oligomerisation, turnover, and localisation.

Different location of adipose tissue may have different consequences to cardiometabolic risk. Here the authors report that deep learning enabled accurate prediction of specific adipose tissue volumes, and that after adjustment for BMI, visceral adiposity was associated with increased risk of cardiometabolic disease, while gluteofemoral adiposity was associated with reduced risk.

DNA methylation is involved in regulatory processes throughout the animal kingdom. Here, the authors map DNA methylation in 535 vertebrates and 45 invertebrates, establishing a reference dataset for cross-species analysis and exploring epigenetic variation across vertebrate evolution.

Developing computational tools for interpretable cell type annotation in scRNA-seq data remains challenging. Here the authors propose a Transformer-based model for interpretable annotation transfer using biologically understandable entities, and demonstrate its performance on large or atlas datasets.

Genomes usually contain multiple chromosomes. The paper reports on GALA, a computational framework for chromosome-based sequencing data separation and gap-free de novo assembly. It allows integration of different sources of data.

Conformational heterogeneity is key to understand how macromolecular function and structure converge. Here, the authors propose an algorithm designed to estimate structural landscapes directly from Cryo-EM particle images.

GPCRs are integral membrane proteins that serve as attractive drug targets. Here, authors delineate the conformational landscapes of 45 GPCRs using a statistical model, highlighting their malleable native ensembles and providing functional insights.

COVID-19-releated public health measures may have indirectly impacted mortality rates by causing or averting deaths. Here, the authors use data from Switzerland until April 2022 and estimate that, after accounting for deaths directly related to COVID-19, mortality was lower than expected, indicating some evidence of an overall positive impact of control measures.

Many software suites and spectral libraries have been developed for DIA proteomics data analysis. Here, the authors create benchmark data sets to evaluate four commonly used software tools combined with seven spectral libraries in both global proteomics and phosphoproteomics analysis.

RNA splicing variations could help identify cancer subtypes, but this task is computationally challenging. Here, the authors develop CHESSBOARD, a Bayesian tile finding algorithm for splicing data which identifies patterns in the form of tiles and can discover leukemia subgroups associated with therapeutic response.

Design of recombinases with new target sites is usually achieved through cycles of directed molecular evolution. Here the authors report Recombinase Generator, RecGen, an algorithm for generation of designer-recombinases; they perform experimental validation to show that this can predict recombinase sequences.

Analysis of a large number of Ribo-seq datasets and genomic alignments led to detection of novel non-AUG proteoforms. Unexpectedly the number of non-AUG proteoforms identified with Ribo-seq greatly exceeds those with strong phylogenetic support.

Longitudinal proteomics holds great promise for biomarker discovery, but the data interpretation has remained a challenge. Here, the authors evaluate several tools to detect longitudinal differential expression in proteomics data and introduce RolDE, a robust reproducibility optimization approach.

The power of pangenomic graphs to improve genetic mapping is still unclear. Here, the authors demonstrate its value in identification of genetic variants associated with disease resistance traits in melon using PanPipes, a pangenome construction and low-coverage genotype-by-sequencing pipeline.

The mutation rate at any specific position in the human genome depends on sequence context. Here, the authors develop a method for predicting mutation rates of point mutations and indels based on sequence context; the results can be used to find genes where de novo mutations cause disease and genes under strong selective constraint.

Covalent labeling (CL) from mass spectrometry experiments provides structural information of higher-order protein structure. Here, the authors develop an algorithm which integrates experimental CL data to predict protein complexes in the Rosetta molecular modeling suite using AlphaFold models.

Previous studies have reconstructed ancestral metabolism using sequence-based approaches. This study uses a high-throughput version of AlphaFold2 to compare proteome-wide 3D structure predictions of two representative strains of ancient archaea and bacteria.

Combining multiple related traits can increase power in genetic association studies. Here, the authors develop a method to integrate GWAS statistics for multiple traits and apply it to find genetic loci affecting human facial variation.

Artificial intelligence prediction accuracy can be reduced with new data. Here, the authors utilise conformal prediction to reduce incorrect predictions in histopathological analysis of prostate cancer biopsies.

Synthetic biology often involves engineering microbial strains to express high-value proteins. Here the authors build deep learning predictors of protein expression from sequence that deliver accurate models with fewer data than previously assumed, helping to lower costs of model-driven strain design.

Observation of the chemical and conformational dynamics of biomolecules by diffraction methods is impeded by several physical artifacts. The authors present an extensible framework for accurate correction of such data that can keep pace with rapid developments in diffraction methods.

Synthetic lethality can be used to identify potential drug targets in cancer based on simultaneous inactivation of two genes through genetic aberrations and gene silencing. Here, the authors develop a statistical framework to identify synthetic lethal pairs from large scale perturbation screens across multiple cancer types.

Single-cell multimodal sequencing technologies are developed to simultaneously profile different modalities of data in the same cell. Here the authors develops a multimodal deep clustering method for the analysis of single-cell multi-omics data that supports clustering different types of multi-omics data and multi-batch data, as well as downstream differential expression analysis.

Inflammatory skin diseases involve various different immune cells in a localised area. Here the authors use spatial transcriptomics to show that disease relevant cytokine transcripts are sparsely expressed in lesional skin, yet are associated with local amplification cascades that promote skin inflammation.

The generation time (interval between successive infections in a transmission chain) is an important parameter for epidemiological modeling. Here, the authors develop a framework for estimating this parameter and how it changes over time and apply it to data from China in the first months of the pandemic.

‘Circulating cell-free DNA can be used to predict cancer, but it is more challenging to assess in early stage cancer. Here, the authors created a diagnostic model using tumor fractions deciphered from circulating cfDNA methylation signatures, which exhibited an 86% sensitivity in detecting early-stage cancer.

Liquid biopsy offers great promise for noninvasive cancer diagnostics, while the lack of adequate target characterization and analysis hinders its wide application. Here, the authors design a transfer learning-based algorithm to transfer lesion labels from the primary cancer cell atlas to circulating tumor cells.

The molecular mechanisms underlying the overlap between oncogenic and embryonic development remain to be explored. Here, the authors use temporal transcriptomic data during development in multiple human organs and suggest the involvement of alternative splicing events, splicing factors, and transcription factors.