Gene co-expression in breast cancer
-| Data Science | Complex Systems | Scientific Research |
During my postdoctoral stay at the Computational Genomics Lab from the National Institute of Genomic Medicine in México City, I worked on developing a hybrid data-driven clustering algorithm that combines eigenvalue decomposition and -medoids to find communities of genes with statistically dependent expression inside each chromosome. We found that the statistical dependency in groups is related to their physical distance in the chromosome and this effect is correlated to the malignancy of the cancer type. This result confirms previous studies of loss of long-range correlation in gene co-expression and contributes to the understanding of this effect in the intra-chromosome scale.
Take a look at the paper here.