Iome and Rifaximin in CirrhosisTable 1. Changes in cognition and cirrhosis severity with rifaximin therapy.citramalic acid after rifaximin. The only significant uni-variate change in urine metabolites was a minor increase in urine succinic acid.N = 20 MELD score INR Serum creatinine (mg/dl) Serum bilirubin (mg/dl) Serum sodium (meq/L) Venous ammonia Cognitive tests Number connection-A (seconds) Number connection-B (seconds) Digit symbol (raw score) Block design (raw score) Line tracing time (seconds) Line tracing errors (number) Serial dotting (seconds)Baseline 9.863.3 1.260.2 0.960.1 1.360.8 138.162.8 46.2623.After rifaximin 9.463.1 1.260.2 0.960.2 1.160.7* 138.962.7 42.9623.Correlation Network AnalysisWe ran the Spearman correlation network analysis on the 2,238 features in the dataset (Table S1) and selected correlation for both 
“Before” and 
“After” treatment that had an absolute Spearman Correlation Coefficient greater than 0.6 and P-value ,0.05 The global correlation networks are very complex with 153,000 correlations (2,220 nodes) for the “before” correlation network (BCN) (MedChemExpress INCB-039110 Figure 4A) and 57,249 correlations (2,225 nodes) for the “after” correlation network (ACN) (Figure 4B). We calculated the intersection correlation network (ICN) which plots all the correlations that are the same in both the BCN and ACN (Figure 4C). Interestingly, over 99 of the features in the dataset are found in the intersection correlation network. Thus, this intersection correlation network delineates the stable core metabiome of the cirrhotic state that didn’t change during treatment. Visually, there is a major hub of urine metabolites with a minor hub of serum metabolites connected by various minor clusters. The complexity of the networks is expected as many compounds will be in the same or complementary metabolic pathway. The networks are visually different and this is reflected in the connectivity measurements (Table 2). For example, the average number of neighbors for the BCN is 59 while it is 51 for the ACN. These parameters indicate that rifaximin has a major effect of the metabolic network, reducing a number of the metabolic interactions and reducing the clustering, while keeping the nodes themselves intact. When we plotted the Cumulative Distribution Function (CDF) of the node degree frequency(14), we found that the connectivity simplified after rifaximin (Figure 4D) and this was a statistically significant shift (P,0.001). We found that most of the nodes included in the BCN and ACN are contained in the ICN [2219 nodes] but it contains a much smaller Thiazole Orange manufacturer subset of the correlations with an average number of neighbors of 13.5. Thus, despite most of the features being present before and after rifaximin therapy, the connectivity changed significantly after rifaximin. This is in contrast to a much more minimal effect on the bacterial abundances of the microbiome. This implies that rifaximin, which is a bacterial RNA polymerase 12926553 inhibitor, does not seem to alter the relative bacterial abundances but does promote a major shift in the complexity of the peripheral metabiome network implying a shift in the gut microbiome functionality. We then calculated the Correlation Difference network (CorrDiff) (Figure 4E) which is a global view of which correlations changed significantly after treatment with rifaximin. We selected only correlation differences that had a Pvalue ,0.05 and where at least one of the original Spearman correlation was greater than 0.6. T.Iome and Rifaximin in CirrhosisTable 1. Changes in cognition and cirrhosis severity with rifaximin therapy.citramalic acid after rifaximin. The only significant uni-variate change in urine metabolites was a minor increase in urine succinic acid.N = 20 MELD score INR Serum creatinine (mg/dl) Serum bilirubin (mg/dl) Serum sodium (meq/L) Venous ammonia Cognitive tests Number connection-A (seconds) Number connection-B (seconds) Digit symbol (raw score) Block design (raw score) Line tracing time (seconds) Line tracing errors (number) Serial dotting (seconds)Baseline 9.863.3 1.260.2 0.960.1 1.360.8 138.162.8 46.2623.After rifaximin 9.463.1 1.260.2 0.960.2 1.160.7* 138.962.7 42.9623.Correlation Network AnalysisWe ran the Spearman correlation network analysis on the 2,238 features in the dataset (Table S1) and selected correlation for both “Before” and “After” treatment that had an absolute Spearman Correlation Coefficient greater than 0.6 and P-value ,0.05 The global correlation networks are very complex with 153,000 correlations (2,220 nodes) for the “before” correlation network (BCN) (Figure 4A) and 57,249 correlations (2,225 nodes) for the “after” correlation network (ACN) (Figure 4B). We calculated the intersection correlation network (ICN) which plots all the correlations that are the same in both the BCN and ACN (Figure 4C). Interestingly, over 99 of the features in the dataset are found in the intersection correlation network. Thus, this intersection correlation network delineates the stable core metabiome of the cirrhotic state that didn’t change during treatment. Visually, there is a major hub of urine metabolites with a minor hub of serum metabolites connected by various minor clusters. The complexity of the networks is expected as many compounds will be in the same or complementary metabolic pathway. The networks are visually different and this is reflected in the connectivity measurements (Table 2). For example, the average number of neighbors for the BCN is 59 while it is 51 for the ACN. These parameters indicate that rifaximin has a major effect of the metabolic network, reducing a number of the metabolic interactions and reducing the clustering, while keeping the nodes themselves intact. When we plotted the Cumulative Distribution Function (CDF) of the node degree frequency(14), we found that the connectivity simplified after rifaximin (Figure 4D) and this was a statistically significant shift (P,0.001). We found that most of the nodes included in the BCN and ACN are contained in the ICN [2219 nodes] but it contains a much smaller subset of the correlations with an average number of neighbors of 13.5. Thus, despite most of the features being present before and after rifaximin therapy, the connectivity changed significantly after rifaximin. This is in contrast to a much more minimal effect on the bacterial abundances of the microbiome. This implies that rifaximin, which is a bacterial RNA polymerase 12926553 inhibitor, does not seem to alter the relative bacterial abundances but does promote a major shift in the complexity of the peripheral metabiome network implying a shift in the gut microbiome functionality. We then calculated the Correlation Difference network (CorrDiff) (Figure 4E) which is a global view of which correlations changed significantly after treatment with rifaximin. We selected only correlation differences that had a Pvalue ,0.05 and where at least one of the original Spearman correlation was greater than 0.6. T.