global metabolic differences between antimonial-sensitive and antimonial-resistant L. donovani isolates

Blue nodes : higher in sensitive strains
Red nodes : higher in resistant strains
node size : fold changes
Node without labels : significant but fold change in lower than 2
P value cutoff < 0.05

http://www.plosntds.org/article/info:doi/10.1371/journal.pntd.0000904

Abstract:
Leishmaniasis is a debilitating disease caused by the parasite Leishmania. There is extensive clinical polymorphism, including variable responsiveness to treatment. We study Leishmania donovani parasites isolated from visceral leishmaniasis patients in Nepal that responded differently to antimonial treatment due to differing intrinsic drug sensitivity of the parasites. Here, we present a proof-of-principle study in which we applied a metabolomics pipeline specifically developed for L. donovani to characterize the global metabolic differences between antimonial-sensitive and antimonial-resistant L. donovani isolates. Clones of drug-sensitive and drug-resistant parasite isolates from clinical samples were cultured in vitro and harvested for metabolomics analysis. The relative abundance of 340 metabolites was determined by ZIC-HILIC chromatography coupled to LTQ-Orbitrap mass spectrometry. Our measurements cover approximately 20% of the predicted core metabolome of Leishmania and additionally detected a large number of lipids. Drug-sensitive and drug-resistant parasites showed distinct metabolic profiles, and unsupervised clustering and principal component analysis clearly distinguished the two phenotypes. For 100 metabolites, the detected intensity differed more than three-fold between the 2 phenotypes. Many of these were in specific areas of lipid metabolism, suggesting that the membrane composition of the drug-resistant parasites is extensively modified. Untargeted metabolomics has been applied on clinical Leishmania isolates to uncover major metabolic differences between drug-sensitive and drug-resistant isolates. The identified major differences provide novel insights into the mechanisms involved in resistance to antimonial drugs, and facilitate investigations using targeted approaches to unravel the key changes mediating drug resistance.




Half of the metabolites could not be mapped in these graphs. Specially, authors have reported 110 lipids species with names but forgot to report any database identifiers. I tried a lot of methods to convert these names into lipidmaps, pubchem or hmdb identifiers but all in vein. It was not possible to convert a name such as GPC (40:6/2) into any identifier. As a result, these lipids species were left out even though they were significantly altered in the study. 


There were also around 50 compound mainly di and tri peptides those could not associated with any CIDs so they could not be included in the network graph. 


It would be very easy to utilize metabolomics datasets for advanced bioinformatics if they were reported with standard compound database identifiers not with confusing and freaking names such as -- (5Z, 7E, 22E, 24E, 26E)-(1S, 3R)-26a, 26b-dihomo-27-nor-9, 10-seco-5, 7, 10(19), 22, 24, 26(26a)-cholestahexaene-1, 3, 26b-triol)