Genetics-driven risk predictions leveraging the Mendelian randomization framework

Daniel Sens; Liubov Shilova; Ludwig Gräf; Maria Grebenshchikova; Bjoern M. Eskofier; Francesco Paolo Casale

PRiMeR is a method that leverages genetic information to learn disease risk predictors across cohorts, circumventing the need for traditional longitudinal studies. With training on risk factors and genetic data from a healthy cohort, along with results from genome-wide association studies (GWAS), PRiMeR can assess risk for new patients. This method was validated on simulations of type 2 diabetes and Alzheimer’s and Parkinson’s disease onset. This method could facilitate more timely and targeted preventive strategies.

Secure discovery of genetic relatives across large-scale and distributed genomic data sets

Matthew Man-Hou Hong; David Froelicher; Ricky Magner; Victoria Popic; Bonnie Berger; Hyunghoon Cho

In this study, Hong et al. developed SF-Relate, a practical and secure federated algorithm for identifying genetic relatives across distributed genomic data sets. Using novel hashing and bucketing strategies, SF-Relate distinguishes relatives from nonrelatives and securely estimates kinship using encrypted data. This method allows for the exclusion of close relatives that can introduce bias in study results while providing privacy protection.

CoRAL accurately resolves extrachromosomal DNA genome structures with long-read sequencing

Kaiyuan Zhu; Matthew Gregory Jones; Jens Luebeck; Xinxin Bu; Hyerim Yi; King L. Huang; Ivy Tsz-Lo Wong; Shu Zhang; Paul S. Mischel; Howard Chang; Vineet Bafna

Circular extrachromosomal DNA (ecDNA) is a form of oncogene amplification found across cancer types and is associated with poor outcome in patients. EcDNAs drive tumor formation, evolution, and drug resistance by modulating oncogene copy number and rewiring gene-regulatory networks. Two methods, CoRAL (Zue et al. 2024) and Decoil (Giurgiu et al. 2024), resolve ecDNA structure using long-read sequencing data, profiling the landscape and evolution of focal amplifications in tumors.

A Bayesian framework for inferring dynamic intercellular interactions from time-series single-cell data

Cameron Y Park; Shouvik Mani; Nicolas Beltran-Velez; Katie Maurer; Teddy Huang; Shuqiang Li; Satyen Gohil; Kenneth J Livak; David A Knowles; Catherine J Wu; Elham Azizi

The method DIISCO characterizes the temporal dynamics of cell–cell interactions in complex biological systems using single-cell RNA sequencing data, elucidating mechanisms underlying normal biological processes and disease progression. This method was demonstrated on simulated and experimental lymphoma–immune interaction data and revealed immune interactions of a cytotoxic T-cell subtype that expands with therapy. This method can guide the design of improved treatments to promote cell states and cross talk crucial for therapeutic response.

Spatial Cellular Networks from omics data with SpaCeNet

Stefan Schrod; Niklas Lück; Robert Lohmayer; Stefan Solbrig; Dennis Völkl; Tina Wipfler; Katherine H. Shutta; Marouen Ben Guebila; Andreas Schäfer; Tim Beißbarth; Helena U. Zacharias; Peter Oefner; John Quackenbush; Michael Altenbuchinger

This paper presents SpaCeNet, a method for analyzing patterns of correlation in spatial transcriptomics data, facilitating reconstruction of both the intracellular and the intercellular interaction networks with single-cell spatial resolution. SpaCeNet was validated on several data sets including mouse visual cortex, mouse organoids, and the Drosophila blastoderm revealing insights into the spatial organization of cells.

Graph-based self-supervised learning for repeat detection in metagenomic assembly

Ali Azizpour; Advait Balaji; Todd J. Treangen; Santiago Segarra

Repetitive DNA poses significant challenges for accurate and efficient genome assembly and sequence alignment. This is particularly true for metagenomic data, where genome dynamics such as horizontal gene transfer, gene duplication, and gene loss/gain complicate accurate genome assembly from microbial communities. Detecting repeats is a crucial first step in overcoming these challenges. Azizpour et al. present GraSSRep, a novel approach that detects and classifies DNA sequences into repetitive and nonrepetitive categories in metagenomics data.