A Computational Biology Approach to Deciphering the Genetic and Cellular Mechani

A Computational Biology Approach to Deciphering the Genetic and Cellular Mechanisms of Aggressive Blood Cancer

by Sagar Sanjiv Desai

Acute Myeloid Leukemia (AML) is a type of blood cancer that has the highest incidence (48.6%) and mortality rate (54%) in Asia. Based on cytogenetic abnormalities and FAB, AML is classified into 8 different subtypes (M0-M7). Defects in granulopoiesis and neutrophil maturation have been linked to mutations in genes associated with neutrophil function. This study focuses on understanding the mutational and expression profiles of transcription factors and proteins involved in neutrophil maturation and degranulation in patients from Caucasian (WES) and Indian populations (WGS). We observed tissue-specific mutation signatures in the blood (ITGAM, XCR1 and CEACAM1) vs bone marrow (HAX1) patient samples. In patients from the Indian population, we observed a missense mutation in HLA-DRB5. Novel deleterious mutations were observed in transcription factors (RUNX1, IKZF1, KMT2A and GATA1) linked to the Regulation of Hematopoiesis. Using CIBERSORTx, we identified variation in the fraction of immune cell types across different subtypes of AML. Monocytes fraction was significantly more in M4-M5 vs M0-M2, whereas Tregs fraction was higher in M0-M2 vs M4-M5. The differential Neutrophil fraction was seen in M0 and M2 vs rest of the subtypes, which correlates with good vs bad prognosis of AML. Comparative analysis of immune cell fractions revealed M0-Macrophage abundant in the bone marrow. Although deficient, bone marrow had a larger representation of neutrophil cell fraction vs blood. To understand differences between neutrophil function in bone marrow vs blood, WGCNA analysis was carried out, and a significant correlation (Pearson correlation coefficient r=0.9, p- value = 3e-18) was observed of genes linked to neutrophil degranulation pathway in the bone marrow. Additionally, gene expression was correlated with FAB classification as a clinical trait revealing subtype-specific HUB genes from various pathways. The gene co-expression analysis was further extended to identify subtype- specific enrichment of lncRNA-mRNA pairs.