Automated classification of depression from structural brain measures across two independent community‐based cohorts

Major depressive disorder (MDD) has been the subject of many neuroimaging case–control classification studies. Although some studies report accuracies ≥80%, most have investigated relatively small samples of clinically‐ascertained, currently symptomatic cases, and did not attempt replication in larger samples. We here first aimed to replicate previously reported classification accuracies in a small, well‐phenotyped community‐based group of current MDD cases with clinical interview‐based diagnoses (from STratifying Resilience and Depression Longitudinally cohort, ‘STRADL’). We performed a set of exploratory predictive classification analyses with measures related to brain morphometry and white matter integrity. We applied three classifier types—SVM, penalised logistic regression or decision tree—either with or without optimisation, and with or without feature selection. We then determined whether similar accuracies could be replicated in a larger independent population‐based sample with self‐reported current depression (UK Biobank cohort). Additional analyses extended to lifetime MDD diagnoses—remitted MDD in STRADL, and lifetime‐experienced MDD in UK Biobank. The highest cross‐validation accuracy (75%) was achieved in the initial current MDD sample with a decision tree classifier and cortical surface area features. The most frequently selected decision tree split variables included surface areas of bilateral caudal anterior cingulate, left lingual gyrus, left superior frontal, right precentral and paracentral regions. High accuracy was not achieved in the larger samples with self‐reported current depression (53.73%), with remitted MDD (57.48%), or with lifetime‐experienced MDD (52.68–60.29%). Our results indicate that high predictive classification accuracies may not immediately translate to larger samples with broader criteria for depression, and may not be robust across different classification approaches.     

Binding mode of conformations and structure-based pharmacophore development for farnesyltransferase inhibitors

Farnesyltransferase (FTase) is one of the prenyltransferase family enzymes that catalyse the transfer of 15-membered isoprenoid (farnesyl) moiety to the cysteine of CAAX motif-containing proteins including Rho and Ras family of G proteins. Inhibitors of FTase act as drugs for cancer, malaria, progeria and other diseases. In the present investigation, we have developed two structure-based pharmacophore models from protein–ligand complex (3E33 and 3E37) obtained from the protein data bank. Molecular dynamics (MD) simulations were performed on the complexes, and different conformers of the same complex were generated. These conformers were undergone protein–ligand interaction fingerprint (PLIF) analysis, and the fingerprint bits have been used for structure-based pharmacophore model development. The PLIF results showed that Lys164, Tyr166, TrpB106 and TyrB361 are the major interacting residues in both the complexes. The RMSD and RMSF analyses on the MD-simulated systems showed that the absence of FPP in the complex 3E37 has significant effect in the conformational changes of the ligands. During this conformational change, some interactions between the protein and the ligands are lost, but regained after some simulations (after 2 ns). The structure-based pharmacophore models showed that the hydrophobic and acceptor contours are predominantly present in the models. The pharmacophore models were validated using reference compounds, which significantly identified as HITs with smaller RMSD values. The developed structure-based pharmacophore models are significant, and the methodology used in this study is novel from the existing methods (the original X-ray crystallographic coordination of the ligands is used for the model building). In our study, along with the original coordination of the ligand, different conformers of the same complex (protein–ligand) are used. It concluded that the developed methodology is significant for the virtual screening of novel molecules on different targets.