Une variante des tests de sphéricité pour l'adéquation de données transcriptomiques à un graphe de régulations génétiques

Vincent Guillemot (1, 2) , Arthur Tenenhaus (2) , Vincent Frouin (1)

1 LEFG - CEA : DSV/iRCM/SCSR/LEFG
2 E3S - Supélec Sciences des Systèmes

When analyzing genomic data, the researcher often encounters the situation where different genetic regulation graphs can be determined on the same dataset with several genetic regulation graph inference methods. One graph is very often compared to another with the help of databases gathering already known interactions between genes: the more known interactions an inferred graph contains, the better it is considered as. We propose a different approach, adapted from the theory of tests, to determine whether a graph fits the dataset.

Keywords: Genomics; graph; hypothesis test; sphericity test.; sphericity test

Domains: Mathematics [math]; Statistics [math.ST]; Statistics [stat]; Statistics Theory [stat.TH]; Engineering Sciences [physics]; Signal and Image processing; Life Sciences [q-bio]; Quantitative Methods [q-bio.QM]; Computer Science [cs]; Bioinformatics [q-bio.QM]; Statistics [stat]; Applications [stat.AP]; Computer Science [cs]; Signal and Image Processing

Vincent Guillemot, Arthur Tenenhaus, Vincent Frouin. Une variante des tests de sphéricité pour l'adéquation de données transcriptomiques à un graphe de régulations génétiques. Journée Ouverte en Biologie et Mathématiques (JOBIM'09), Jun 2009, Nantes, France. pp.219-220. ⟨hal-00421943⟩