Fig. 9: The importance of derived predictive biomarkers with each method for Humira HS studies.
From: Deep learning-based ranking method for subgroup and predictive biomarker identification in patients
a DeepRAB. b XGboostML. c CF. d LRMC. XGboostML: XGboost with modified loss function; CF: causal forest; LRMC: linear regression with modified covariates; CHG_AN: % reduction in AN count at week 12; AN12: AN count at week 12; AN0: AN count at week 0; DFT_OBS_0: draining fistula count at week 0; DFT_OBS_12: draining fistula count at week 12; CHG_DFT_OBS: reduction in draining fistula count at week 12; ABT_OBS_0: abscess count at week 0; ABT_OBS_12: abscess count at week 12; CHG_ABT_OBS: reduction in abscess count at week 12; SMOKER: smoking status; STHUSTN: Hurley stage at week 0; ITT2RFN: initial responder status at week 12; HISCR0: HiSCR at week 0; STANHSN: concomitant use of antibiotics.
