Objective To screen and verify the rheumatoid arthritis(RA)-related ferroptosis characteristic genes by bioinformatics methods,and predict the traditional Chinese medicine that could act on these feature genes.Methods Microarray data of joint synovial tissue of RA and healthy controls were downloaded from Gene Expression Omnibus(GEO)database,and the RA-related ferroptosis dif-ferential expression gene(FDEG)were obtained through the FerrDb database,and the biological connotations and signaling pathways of FDEG were explored through Gene Ontolog(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)analysis.Machine learning was used to obtain the feature FDEG,and the diagnostic ability was judged by combining it with the external validation set analysis.The symMap database was used to reverse predict the traditional Chinese medicine that could act on these feature genes.Results A total of 484genes related to ferroptosis were downloaded and 125 FDEG with significant differences were selected from the list of genes in the train-ing set.GO enrichment analysis revealed that these 125 FDEG could be co-morbid from multiple biological perspectives.KEGG analysis revealed their roles in ferroptosis,autophagy,tumors,and hypoxia-inducible factor(HIF-1),forkhead box protein O1(FoxO1),in-terleukin-17(IL-17),and tumor necrosis factor(TNF)signaling pathways.Machine learning obtained three RA ferroptosis character-istic genes,such as ATF3,AKR1C1 and RRM2.After analyzing the expression levels of the ferroptosis characteristic genes in the training and validation sets,it was found that AKR1C1 and RRM2had certain diagnostic efficacy in the RA Ferroptosis pathway.Then,gene set enrichment analysis(GSEA)and gene set variation analysis(GSVA)were performed on the ferroptosis characteristic genes,and it was found that AKR1C1,ATF3,and RRM2 were mainly closely related to metabolism,synthesis,and other biological processes.Finally,the symMap database predicted that the three ferroptosis characteristic genes could be affected by traditional Chinese medicines,and it was found that the phlegm-reducing and qi-regulating traditional Chinese medicine accounted for the largest proportion of the total.Conclu-sion AKR1C1 and RRM2genes have certain diagnostic efficacy in the RA ferroptosis pathway,and qi-regulating and phlegm-reduc-ing traditional Chinese medicine can be used as a reference for the anti-RA ferroptosis pathway.

Objective To screen and verify the rheumatoid arthritis(RA)-related ferroptosis characteristic genes by bioinformatics methods,and predict the traditional Chinese medicine that could act on these feature genes.Methods Microarray data of joint synovial tissue of RA and healthy controls were downloaded from Gene Expression Omnibus(GEO)database,and the RA-related ferroptosis dif-ferential expression gene(FDEG)were obtained through the FerrDb database,and the biological connotations and signaling pathways of FDEG were explored through Gene Ontolog(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)analysis.Machine learning was used to obtain the feature FDEG,and the diagnostic ability was judged by combining it with the external validation set analysis.The symMap database was used to reverse predict the traditional Chinese medicine that could act on these feature genes.Results A total of 484genes related to ferroptosis were downloaded and 125 FDEG with significant differences were selected from the list of genes in the train-ing set.GO enrichment analysis revealed that these 125 FDEG could be co-morbid from multiple biological perspectives.KEGG analysis revealed their roles in ferroptosis,autophagy,tumors,and hypoxia-inducible factor(HIF-1),forkhead box protein O1(FoxO1),in-terleukin-17(IL-17),and tumor necrosis factor(TNF)signaling pathways.Machine learning obtained three RA ferroptosis character-istic genes,such as ATF3,AKR1C1 and RRM2.After analyzing the expression levels of the ferroptosis characteristic genes in the training and validation sets,it was found that AKR1C1 and RRM2had certain diagnostic efficacy in the RA Ferroptosis pathway.Then,gene set enrichment analysis(GSEA)and gene set variation analysis(GSVA)were performed on the ferroptosis characteristic genes,and it was found that AKR1C1,ATF3,and RRM2 were mainly closely related to metabolism,synthesis,and other biological processes.Finally,the symMap database predicted that the three ferroptosis characteristic genes could be affected by traditional Chinese medicines,and it was found that the phlegm-reducing and qi-regulating traditional Chinese medicine accounted for the largest proportion of the total.Conclu-sion AKR1C1 and RRM2genes have certain diagnostic efficacy in the RA ferroptosis pathway,and qi-regulating and phlegm-reduc-ing traditional Chinese medicine can be used as a reference for the anti-RA ferroptosis pathway.