Jan 26 - Feb 1, 2026
Spatial Multi-Omics, Single-Cell Perturbations, and Variant Classification
February 7, 2026
Can we profile multiple molecular layers in a single tissue section? How accurately can we predict cellular responses to genetic perturbations? Which variant classification tool should you trust for clinical diagnosis? This week explores answers through spatial multi-omics, perturbation prediction benchmarks, and more.
Spatial Multi-Omics Integration
A new method enables high-parameter spatial multi-omics through histology-anchored integration, allowing researchers to simultaneously profile multiple molecular layers within intact tissue sections. This advancement bridges the gap between imaging and sequencing, enabling cell-type-specific analysis of gene expression, chromatin accessibility, and protein localization within the same tissue context. The approach significantly enhances understanding of tissue architecture and cell-cell interactions in development and disease.
Single-Cell Perturbation Response Prediction
A comprehensive benchmark evaluates algorithms for generalizable single-cell perturbation response prediction. The study compares various computational approaches for predicting how individual cells respond to genetic or chemical perturbations, revealing key differences in predictive accuracy and generalization capability. This resource helps researchers select appropriate methods for drug response prediction and functional genomics screens.
Variant Classification Tools Evaluation
A systematic comparison evaluates ACMG/AMP-based variant classification tools, assessing their accuracy and consistency in interpreting genetic variants. The study reveals significant variability between tools, highlighting the need for standardized approaches in clinical variant interpretation. These findings have important implications for diagnostic laboratories and precision medicine applications.
Integrative Multi-Omics Module Analysis
A novel framework called iModMix enables integrative module analysis across multiple omics datasets, allowing researchers to identify coordinated functional modules spanning different molecular layers. The method helps uncover relationships between genetic variants, gene expression, proteomics, and other data types, facilitating systems-level understanding of biological processes and disease mechanisms.
Trajectory-Based Protein pKa Prediction
A new tool called TrIPP provides trajectory-based pKa prediction for protein engineering applications. The method predicts ionization states of amino acid residues along protein dynamics trajectories, enabling more accurate modeling of protein function and stability. This advancement supports rational design of engineered proteins for therapeutic and industrial applications.