Thejus Jayakrishnan: Host-microbiome interactions in early vs average-onset CRC
Thejus Jayakrishnan, Hematology/Oncology Fellow at Cleveland Clinic, shared a post on X:
”Publication alert.
Multi-omics machine learning to study host-microbiome interactions in early-onset colorectal cancer.
Authors: Thejus T. Jayakrishnan, Naseer Sangwan, Shimoli V. Barot, Nicole Farha, Arshiya Mariam, Shao Xiang, Federico Aucejo, Madison Conces, Kanika G. Nair, Smitha S. Krishnamurthi, Stephanie L. Schmit, David Liska, Daniel M. Rotroff, Alok A. Khorana, Suneel D. Kamath.
We use ‘multiomics machine learning to study host-microbiome interactions in early-onset vs average-onset colorectal cancer.’
Distinct clustering patterns were observed in the multi-omic dimension reduction plots using metabolomics vs microbiome.
The metabolomics classifier achieved an AUC of 0.98 vs AUC 0.61 for microbiome-based classifier. Several unique metabolites and microbiome features contributed to the models.
Circular correlation technique highlighted several key associations. Eg: Metabolites glycerol and pseudouridine (higher abundance in individuals with aoCRC) had negative correlations with Parasutterella, and Ruminococcaceae (higher abundance in individuals with eoCRC).
Network analysis, avoiding preselection bias in the prior models showed: Metabolites of the urea cycle, including urea and uric acid; Citric acid; Microbe Akkermansia showed different clustering patterns and centrality.
Conclusions:
- Multiomics analysis can be utilized to study host-microbiome correlations in eoCRC and aoCRC.
- Metabolomics show promising biomarker potential.
- Distinct host-microbiome correlations for urea cycle therapeutic opportunity.
Huge thanks to my mentors, coauthors, and individuals who participated in the study – Shimoli V Barot, Nicole Farha, Federico Aucejo, Madison Conces, Smitha Krishnamurthi, Stephanie Schmit, David Liska, Daniel Rotroff, Arshiya Mariam, Shao Xiang, Cleveland Clinic.”
Source: Thejus Jayakrishnan/X