Fig. 1: Multi-omics profiles of IBS patients.

a Principal coordinates analysis (PCoA) of discovery cohort individuals based on 330 gut metagenomic (264 IBS and 66 HC), 330 fecal metabolomic (264 IBS and 66 HC), and 325 serum metabolomic profiles (259 IBS and 66 HC), respectively (Bray−Curtis distance). b Left panel shows the comparisons of gut microbiota dysbiosis in different diseases and different subtypes of IBS. The gut metagenomic data of other diseases, including inflammatory bowel disease (IBD, n = 155, including 121 IBD and 34 HC) [25], liver cirrhosis (LC, n = 237, including 123 LC and 114 HC) [26], colorectal cancer (CRC, n = 128, including 74 CRC and 54 HC) [27], and type 2 diabetes (T2D, n = 363, including 178 T2D and 185 HC) [28], were collected by the R package curatedMetagenomicData [29]. Right panel shows the comparisons of gut microbiota dysbiosis in different subtypes of IBS, including IBS-C (n = 90, 24 IBS-C and 66 HC), IBS-D (n = 280, 214 IBS-D, and 66 HC), IBS-M (n = 85, 19 IBS-M, and 66 HC), and IBS-U (n = 73, 7 IBS-U, and 66 HC). c Shown are associated factors with gut metagenomic, fecal metabolomic (negative ionization mode), and serum metabolomic (negative ionization mode) profiles in the 330-member discovery cohort. Black bars indicate statistical significance (FDR < 0.1).