2026

Molecular Communication Mechanisms Underlying Inflammation-Driven Fibrosis

CSIBD investigators have identified key molecular mechanisms that drive fibrosis in IBD. Fibrosis is a complication of conditions like Crohn’s disease and ulcerative colitis that can lead to organ dysfunction. In a study published in Nature, the team mapped individual cells from patients’ intestinal tissue and found that interactions between inflammatory macrophages and inflammation-associated fibroblasts promote scar tissue formation through production of IL-11, a protein that deposits scar tissue. They also discovered that GLIS3, a transcription factor not previously linked to IBD, is a central regulator of this cellular communication loop. Experimental results showed that blocking GLIS3 prevents fibrosis in in vivo models, suggesting that targeting the GLIS3-IL-11 pathway could lead to new treatments for IBD and other chronic inflammatory diseases marked by fibrosis.

Read the study

Scaling Microbiome Insight: New Methods for Association Testing, Function Prediction, and Virome Profiling

The Huttenhower lab has expanded their computational toolbox for comprehensive microbiome profiling and analysis:

  • Microbiome Multivariable Associations with Linear Models (MaAsLin 3) identifies microbial features linked to environmental or health phenotypes by separately modeling abundance and prevalence associations in microbiome multi-omics data. Applied to pediatric and adult IBD cohorts, MaAsLin 3 recapitulated known associations and uncovered new prevalence-based associations, improving the accuracy and specificity of detected relationships. Read the study.
  • Function predictor of Uncharacterized Gene products by Assessing high-dimensional community data in Microbiomes (FUGAsseM) predicts functions for previously uncharacterized microbial proteins by integrating metatranscriptomic coexpression with genomic proximity, sequence similarity, and domain interaction evidence. FUGAsseM generated high-confidence functional predictions for more than 443,000 protein families, including many with little or no homology to known proteins, enabling systematic study of undercharacterized microbial communities. Read the study.
  • Bioinformatic Application for Quantification and Labeling of Viral Taxonomy (BAQLaVa) profiles viral communities from metagenomic and metatranscriptomic data, enabling detection of more than 120,000 viral species. Applied to the HMP2 IBDMDB cohort, BAQLaVa identified virome perturbations in IBD and inferred phage–host relationships without reference genomes, representing a scalable framework for virome epidemiology and systematic analysis of