Deconvolution of rat pancreatic islet RNA-seq to uncouple islet cell composition effect from INGAP peptide treatment

HEIDENREICH, ANA C.; ROMERO, AGUSTÍN; ROMAN, CAROLINA L.; GAGLIARDINO, JUAN J.; MAIZTEGUI, BÁRBARA; FLORES, LUIS E.; DI PERSIA, LEANDRO; MILONE, DIEGO H.; RODRÍGUEZ SEGUÍ, S.A.

Encuentro en formato virtual.

Encuentro; 1st Women in Bioinformatics & Data Science Latin America (WBDS-LA); 2020

Women in Bioinformatics & Data Science Latin America

Diabetes Mellitus is a disease characterized by the loss or reductionof the pancreatic β cell mass, with the consequent impairment ofinsulin production. A pentadecapeptide derived from the IsletNeogenesis Associated Protein (INGAP-PP) has been previously shown toincrease the β cell mass and insulin secretion in normal anddiabetic animals (Kapur, 2012). In this work, we performedcuadriplicate RNA-seq assays on rat pancreatic islets treated invitro with INGAP-PP to gain insights into the mechanismsmodulated by the peptide treatment.Sequence read alignment was performed using HiSAT, followed byStringtie for de novo gene annotation, transcript expressionquantification and normalization across samples. Pancreatic isletsare a heterogeneous tissue, composed of at least fourteen cell typesincluding exocrine, endocrine, vascular and immune cells, amongothers. As well, the INGAP-PP can induce transcriptional changes onseveral of these cell types. Indeed, a standard RNA-seq analysispipeline did not allow a clear discrimination of the INGAP-PPtreatment transcriptional effects. By combining our bulk RNA-seqquantification with gene expression profiles obtained from isletsingle cell RNA-seq data, different levels of expression were trackedin samples (gene subsets associated with specific islet cellcomponents in control RNA-seq). This finding lead us to developbioinformatics tools tomodel and deconvolvethe original islet cell composition in control samples.Deconvolution process is implemented through an iterative algorithmwith factorization of non-negative matrices. Therefore, theproportion of each cell type of the samples can be inferred, thusallowing a better determination of the differential expression. Ourongoing work aims at using these tools to uncouple the islet cellcomposition effect from the INGAP-PP treatment, ultimately allowingfor a more precise analysis of thesignaling pathwaysmodulated by the peptide treatment. p { margin-bottom: 0.1in; line-height: 115% }a:link { so-language: zxx }