Benchmark research code#

All analyses are reproducible using R code freely available on Github. Contact: +49 89 2180 3198 Supplementary information: Supplementary data are available at Briefings in Bioinformatics online. This indicates, as a by-product of our study, that in the considered TCGA studies the utility of multi-omics data for prediction purposes was limited. Moreover, only the block forest method outperformed the Cox model on average, and only slightly. Taking this structure into account can protect the predictive information in low-dimensional groups-especially clinical variables-from not being exploited during prediction. The results indicate that methods taking into account the multi-omics structure have a slightly better prediction performance.

UnoвЂ™s C-index and the integrated Brier score served as performance metrics. The methods were compared using several repetitions of 5-fold cross-validation. The KaplanвЂ“Meier estimate and a Cox model using only clinical variables were used as reference methods. Eleven methods based on boosting, penalized regression and random forest were compared, comprising both methods that do and that do not take the group structure of the omics variables into account. The considered outcome was the (censored) survival time. Different prediction methods from machine learning and statistics were applied on 18 multi-omics cancer datasets (35 to 1000 observations, up to 100 000 variables) from the database вЂThe Cancer Genome AtlasвЂ™ (TCGA).

We aim to give some answers to these questions through a large-scale benchmark study using real data. It is also unclear which methods are most appropriate to derive such prediction models. Nevertheless, questions remain regarding the usefulness of multi-omics data for the prediction of disease outcomes such as survival time. Multi-omics data, that is, datasets containing different types of high-dimensional molecular variables, are increasingly often generated for the investigation of various diseases.

Benchmark research