Are there particular microbial strains of interest that your working with and have access to? Are there specific metabolites that you are looking to detect? Simply collecting a complete metabolite profile (by LCMS) of only the serum or plasma would not tell you much by itself. Detecting microbial metabolites in serum or plasma can be difficult even when you know exactly what you are looking for.

A general approach would be to analyze the microbial metabolites (extracted from culture or using reference standards) and also the serum/plasma by HPLC-MS in sequence,. Then you could compare the LCMS data to find mass/retention time matches. For the MS part of this, you really want at least high resolution MS (HRMS) since unit or low resolution MS often produces coincidental false "matches". Fast polarity switching (+/-) and data-dependent MS/MS would also be preferred for the MS (ex. Shimadzu IT-TOF). 

Another challenge is that, since serum and plasma are very complex matrices, the signal for many of the microbial metabolites may be obscured or suppressed.  For example, some metabolites may be bound to proteins in the serum or plasma. Others may be metabolically transformed. You would need to do sample cleanup on the serum and plasma (such as desalting, protein precipitation, or solid-phase extraction).

If you know exactly which specific microbial metabolites you are looking for and have access to these compounds, then this becomes much easier. You can use a HPLC - triple quadrupole MS to tune for, detect, and even quantify specific microbial metabolites in the serum or plasma. 

Good Luck!

Dear Andrew, Thank you for your answer, it has definitely helped me out in more detailed planning!

However, for now, I don't know the specific metabolites, and I am trying to look for metabolites which are differentially present in two groups of samples, and one of the sub group of metabolites I was interested in looking for, was the bacterial metabolites hence the question!

Thank you again !

Hi, it depends on what kind of bacterial metabolites do you want to detect on human plasma, if you know the sort of bacteria you can detect the metabolites by using specific antibodies. 

If you could get a metabolite profile of the microbe(s) of interest, then you have a high chance of successfully identifying some of those metabolites in the plasma/serum. You could acquire an LCMS profile of the microbial extract(s) along with the biological samples and match peaks. 

if the bacterial class occurs in one individual and not in another you can do a comparative metabolomic investigation of the two sera using LC-MS data and multivariate stats