Sample size and genetic architecture complexity constrain deep learning benefits

Demonstrates that the performance and advantages of deep learning models depend strongly on sample size and the complexity of genetic architecture present in the data, establishing these as key mediators.

Part of Chain