Structural analysis of human ATE1 isoforms and their interactions with Arg-tRNAArg.

Posttranslational protein arginylation has been shown as a key regulator of cellular processes in eukaryotes by affecting protein stability, function, and interaction with macromolecules. Thus, the enzyme Arginyltransferase and its targets, are of immense interest to modulate cellular processes in the normal and diseased state. While the study on the effect of this posttranslational modification in mammalian systems gained momentum in the recent times, the detail structures of human ATE1 (hATE1) enzymes has not been investigated so far. Thus, the purpose of this study was to predict the overall structure and the structure function relationship of hATE1 enzyme and its four isoforms. The structure of four ATE1 isoforms were modelled and were docked with 3'end of the Arg-tRNAArg which acts as arginine donor in the arginylation reaction, followed by MD simulation. All the isoforms showed two distinct domains. A compact domain and a somewhat flexible domain as observed in the RMSF plot. A distinct similarity in the overall structure and interacting residues were observed between hATE1-1 and X4 compared to hATE1-2 and 5. While the putative active sites of all the hATE1 isoforms were located at the same pocket, differences were observed in the active site residues across hATE1 isoforms suggesting different substrate specificity. Mining of nsSNPs showed several nsSNPs including cancer associated SNPs with deleterious consequences on hATE1 structure and function. Thus, the current study for the first time shows the structural differences in the mammalian ATE1 isoforms and their possible implications in the function of these proteins.Communicated by Ramaswamy H. Sarma.

HPCGCN: A Predictive Framework on High Performance Computing Cluster Log Data Using Graph Convolutional Networks.

This paper presents a novel use case of Graph Convolutional Network (GCN) learning representations for predictive data mining, specifically from user/task data in the domain of high-performance computing (HPC). It outlines an approach based on a coalesced data set: logs from the Slurm workload manager, joined with user experience survey data from computational cluster users. We introduce a new method of constructing a heterogeneous unweighted HPC graph consisting of multiple typed nodes after revealing the manifold relations between the nodes. The GCN structure used here supports two tasks: i) determining whether a job will complete or fail and ii) predicting memory and CPU requirements by training the GCN semi-supervised classification model and regression models on the generated graph. The graph is partitioned into partitions using graph clustering. We conducted classification and regression experiments using the proposed framework on our HPC log dataset and evaluated predictions by our trained models against baselines using test_score, F1-score, precision, recall for classification, and R1 score for regression, showing that our framework achieves significant improvements.