Bioinformatic evaluation of differentially expressed genes in Myelodysplastic Syndromes with SF3B1 variants

LINCANGO YUPANKI, MARCO; LARRIPA, IRENE; BELLI, CAROLINA B

Mar del Plata

Congreso; Reunión Conjunta SAIC, SAI, FAIC, SAFIS; 2022

SAIC

Bioinformatic analysis of cloud databases is an emerging research method to dig the physiopathology of diseases. Pathogenic variants in the splicing factor (SF) SF3B1 (SF3B1mut) in low risk Myelodysplastic Syndromes (MDS) were recognized within a distinct subtype in 2022, associated with ring sideroblasts, good outcome and response to Luspatercept. Our aim was to individualize differential expressed genes (DEG) related with SF3B1mut and a differential outcome in MDS.RNAseq datasets of CD34+ of MDS patients (24 SF3B1mut, 32 SF-wild type-WT) and healthy controls (7 HC) were downloaded from the GEO database GSE114922. Analysis, specified below, were performed in R v4.2.1.A total of 1342 DEG (p0.85) to determine DEG related with the presence SF3B1mut highlighted MRPL2 (0.87), MRPL52 (0.86), NDUFA8 (0.89), MRRF (0.90), ALAD (0.85) and GAR1 (0.91). SF3B1mut was independently associated (Logistic regression) with higher expression levels of MRRF (OR36.3, CI95% 1.6-1827.4, p=0.04) and a tendency for NDUFA8 (OR25.7, CI95% 0.8-1545.9, p=0.08) and GAR1 (OR 23.9, CI95% 0.8-3656.9, p=0.14). Regarding survival (Cox’s regression), higher expression levels of NDUFA8 (HR0.2, CI95% 0.04-0.8, p=0.024) and of GAR1 (HR 0.1, CI95% 0.02-0.7, p=0.022) were independently associated with a better outcome, while of MRPL52 (HR 7.9, CI95% 0.8-79.1, p=0.08) and MRRF (HR 21.5, CI95% 0.6-753.9, p=0.091) with a tendency of a worse one.Our in silico approach, which validation is ongoing, underlined four DEG associated with differential survival outcome and the presence of SF3B1 variants in MDS.