Figure 5
From: Benchmarking protein language models for protein crystallization
Comparison of area under precision-recall curve (AUPR) of benchmark PLMs for the crystallization prediction task. (a) AUPR for fairly balanced test set using XGBoost, (b) AUPR for SP_final dataset using XGBoost, (c) AUPR for TR_final dataset using XGBoost, (d) AUPR for fairly balanced test set using LightGBM, (e) AUPR for SP_final dataset using LightGBM, and (f) AUPR for TR_final dataset using LightGBM.
