Authors: Murillo Javier; Spetale Flavio Ezequiel; Garcia Labari Ignacio; Tapia Elizabeth; Bulacio Pilar.
Resumen: Many Bioinformatics tools have been developed to predict the effect of single nucleotide polymorphisms (SNPs) on gene functionality in an effort to reduce the need for in-vivo assays. Their use in scientific research and even in personalized medicine is frequent and the need for understanding their output, essential. Many tools with heterogeneous outputs are available. This makes their selection, understanding, and comparison a non-trivial task. Current methods comparing these tools overcome the heterogeneity of outputs problem by binarizing them at the expense of information loose. We are concerned about the consequences of these practices. In this work, we pose two closely related questions as a starting point: i) Do these prediction tools provide similar results -inner consistency?, and ii) Are their predictions consistent with the literature results -outer consistency?To answer these questions, 6 prediction tools were selected based on the diversity of underlying learning methods, on the possibility of their online evaluation, and on their popularity in literature. Two indices, of different penalization strength, were proposed to evaluate their outer and inner consistencies. Proposed indices quantify the systematicdisagreement between each pair of SNPs, i.e., count pairs of predictions ordered differently in each tool scale, without performing any scale normalization. The tools were tested with 2730 SNPs from breast cancer type one susceptibility protein encoded by the BRCA1 gene.
Meeting type: Conferencia.
Type of job: Artículo Breve.
Production: On the consistency of tools that predict the impact of SNPs in the BRCA1 gene.
Scientific meeting: A2B2C 10th Meeting.
Meeting place: Mendoza.
Organizing Institution: Universidad Nacional de Cuyo.
It's published?: Yes
Publication place: Mendoza
Meeting month: 11