nsalean.blogg.se

Differences from the reference genome were identified using the CaVEMan and Pindel algorithms identifying substitution and small insertions/deletions, respectively ( ). A matched lympoblastoid line was used as the reference genome for a small subset (n=39) of the cell lines where this was available. Variants were identified by comparison to a reference genome. Validation of the ANOVA interactions and LOBICO models on independent datasets Model selection, filtering and projection on primary tumor data LOBICO: plotting a selected group of AND and OR combinations LOBICO: finding interesting pairwise AND and OR combinations LOBICO: statistical test to identify predictive models LOBICO: cross-validation and feature importance scores Pathway activity signatures and inference (SPEED analysis) Tumors/cell lines integrated analysis: enrichment analysis of individual cancer functional event occurrences in a given classĬell viability assay, dose response curve fitting model, comparison across replicates and compound cluster analysisĪNOVA model, effect-size, significance and p-value correction Tumors/cell lines integrated analysis: enrichment analysis of event types across the identified classes Tumors/cell lines integrated analysis: identification of classes of samples and selected functional events Tumors/cell lines integrated analysis: Knn classification performances and euclidean embedding Tumors/cell lines integrated analysis: Frequency profile comparisons Transcriptional data generation and pre-processing Identification of informative CpG island and methylation data discretisation Identification of recurrently copy number altered chromosomal regions