Causal network inference based on cross-validation predictability

Identifying causal relations or causal networks among molecules/genes, rather than just their correlations, is of great importance but challenging in biology and medical field, which is essential for unraveling molecular mechanisms of disease progression and developing effective therapies for disease treatment. However, there is still a lack of high-quality causal inference algorithms for any observed data in contrast to time-series data. In this study, we developed a causal concept for any observed data based on cross-validated predictability (CVP). The CVP can quantify the causal effects among observed variables in a system. The causality was extensively validated by combining a large variety of statistical simulation experiments and available benchmark data (simulated data and various real data). Combining the predicted causal network and the real benchmark network, the CVP algorithm demonstrates high accuracy and strong robustness in comparison with the mainstream algorithms. Identifying causal relations or causal networks among molecules/genes is of great importance but challenging. By introducing a causal inference algorithm based on cross-validated predictability (CVP), the research demonstrates high accuracy and robustness in predicting causal networks independent of time-series data and directed acyclic graph (DAG)