The basic features of some of the most versatile and popular open source frameworks for machine learning (TensorFlow, Deep Learning4j, and H2O) are considered and compared. Their comparative analysis was performed and conclusions were made as to the advantages and disadvantages of these platforms. The performance tests for the de facto standard MNIST data set were carried out on H2O framework for deep learning algorithms designed for CPU and GPU platforms for single-threaded and multithreaded modes of operation.

Existing molecular modeling software and different techniques of molecular dynamics trajectories analysis were evaluated in terms of subsequent integration into the MolDynGrid virtual laboratory services. GROMACS 4 molecular dynamics package contains the most comprehensive set of analytical programs, thus the standard analytical tools in MolDynGrid were based on this software. G_correlation tool was ported to GPU which dramatically increased its productivity. Scripting capabilities of VMD software were used for certain non-standard analysis techniques, while complex and computationally intensive techniques were implemented using Pteros molecular modeling library. Parallel execution of analytical programs was implemented in MolDynGrid by a set of internally developed scripts.