The metabolite list from the study by Srinivasan et. al. (2015) reported 108 metabolites with their KEGG annotations and was used for our comparison (28) . From the MMIP predicted list of significant metabolites, 52 overlapped with their list and 21 of those 52 metabolites depicted the same trend as reported in the study. We provide a relative value that signifies the capacity of that community to consume or accumulate a particular metabolite. Like for example cadaverine, in the study they showed that the BV+ samples had a higher amount of this compound; in our prediction we showed that compared to that of BV-(or H) group, the relative capacity of cadaverine being accumulated in the BV+ samples is significantly higher than that of BVsamples. With regards to amino acids, we correctly predicted the lowering of amino acids in BV+ samples. The study reported that 18 out 19 amino acids that they found in the samples, were lower in amount in BV+ samples compared to that of BV-or healthy samples; there is an increase in amino acid catabolites in BV+ samples. The amino acids that lowered in our prediction included alanine, cysteine, histidine, tryptophan, phenylalanine etc. and the dipeptide, alanylalanine. Also, the compounds that are involved in amino acid metabolism pathways like spermine and glutathione, were also found to be low in BV+ and matched with that of the study. While drawing a biological inference from this result we can say, an increase in the number of bacterial taxa and not just Lactobacillus, probably cause an increased usage of amino acids since they are essential during cell growth; correlating to this pattern we can also say that this should also increase the amount of amino acid catabolites, and this too is confirmed by our prediction. Thus, the increase in amino acid catabolites and decrease of amino acids in BV+ samples with increased bacterial diversity.
Focusing on the carbohydrate metabolism associated metabolites, the study reported lowering of simple sugars in BV+ environment, and the same is observed in our results too; maltose being one of them, that overlap with them and follow the correct trend. Also, glucose-6-phosphate, and sugar alcohols like sorbitol and mannitol are other reported carbohydrate metabolism metabolites that are correctly predicted and maintain the trend of being present in lower amounts in BV+ cases. The same explanation that justified our amino acids metabolite prediction could also justify the case for carbohydrate metabolism metabolites; while the simple sugars and intermediate metabolites are being used up, the carbohydrate catabolites tend to accumulate in BV+ samples with the growing diversity of microbes.
They also reported lowering of glycerol, ethanolamine and glycerophosphorylcholine (GPC), all of which are associated with lipid metabolism, in BV+ samples and they too were correctly predicted.
Focusing on the carbohydrate metabolism associated metabolites, the study reported lowering of simple sugars in BV+ environment, and the same is observed in our results too; maltose being one of them, that overlap with them and follow the correct trend. Also, glucose-6-phosphate, and sugar alcohols like sorbitol and mannitol are other reported carbohydrate metabolism metabolites that are correctly predicted and maintain the trend of being present in lower amounts in BV+ cases. The same explanation that justified our amino acids metabolite prediction could also justify the case for carbohydrate metabolism metabolites; while the simple sugars and intermediate metabolites are being used up, the carbohydrate catabolites tend to accumulate in BV+ samples with the growing diversity of microbes.
They also reported lowering of glycerol, ethanolamine and glycerophosphorylcholine (GPC), all of which are associated with lipid metabolism, in BV+ samples and they too were correctly predicted.
