Drug
repositioning is the process of recycling ex- isting drugs for new indications
by identifying the potential drug-target interactions (DTIs). However, predicting new associations
between drugs and target proteins is a challenging issue because the number of
known interacting drug-target pairs is much smaller than that of
non-interacting drug-target pairs which have not been experimentally verified
to be true negative samples. Moreover, the
explosive growth in genomic sequence and chemical structure data motivate us to
develop an effective computational method to accurately predict a novel
drug-target interaction. In this paper, we
propose an algorithmic framework based on deep semi supervised learning called
DSSL-DTIs to potentially overcome these limitations. Firstly, we use the
powerful technique of stacked autoencoder to convert high- dimensional features
to low-dimensional representations. Then, we apply unsupervised stacked
autoencoder model for initializing
the weights of a supervised deep neural networks model. The proposed approach
has been compared
with other state-of-the-art methods applied all on
the same reference datasets of Drug-Bank.

Preliminary performance assessment results have shown that our approach
outperforms these techniques. Its overall accuracy performance is more than
98%. The DSSL-DTIs can be used     to
predict large-scale new drug-target interactions. The obtained results have
shown that highly ranked candidate DTIs obtained from DeepSS-DTIs are also
present in the DrugBank database and in the literature which demonstrates the
effectiveness of the proposed approach.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now