Search bioRxiv⌕ Search

Biology subjects

Ghosh, Z.

Publications and source records attributed to Ghosh, Z..

2 recordsLinked to original sources

LncPTPred: Predicting lncRNA-Protein Interaction based on Crosslinking and Immunoprecipitation (CLIP-Seq) Data

Long non-coding RNA (lncRNA)-Protein Interaction (LPI) across diverse biological systems directly and indirectly regulates various cellular processes. Experimental assays to recognize the protein binding partners of lncRNAs is highly time consuming and expensive. In-silico predictive approaches involving pattern recognition technique provides a promising alternative to it by reducing the search space. Our work identifies such hidden pattern from within the Cross-linking immunoprecipitation sequencing (CLIP-Seq) data which aids to overcome the problem of obtaining real negative dataset and thus offers a state-of-the-art machine learning (ML) based prediction algorithm to predict LPI. Initial phase of this work involves preparation of the training dataset and the next phase is devoted towards developing the ML based model to perform prediction operation. In order to show the efficacy of our model, its performance has been compared against that of the contemporary prediction tools, where the result clearly shows the outperformance of our model. Moreover it also provides the segments of interaction within the lncRNA loci which acts as a roadmap for precise designing of the validation experiment. The LncPTPred tool has been provided in terms of web server as well as standalone version in github. Web server Link: http://bicresources.jcbose.ac.in/zhumur/lncptpred/ Github Link: https://github.com/zglabDIB/lncptpred.git

bioinformatics↗

SecDATA: Secure Data Access and de novo TranscriptAssembly protocol - To meet the challenge of reliableNGS data analysis

Recent developments in sequencing technologies have created new opportunities to generate high-throughput biological data at an affordable price. Such high-throughput data needs immense computational resources for performing transcript assembly. Further, a high-end storage facility is needed to store the analyzed data and raw data. Here comes the need for centralized repositories to store such mountains of raw and analyzed data. Hence, it is of utmost importance to ensure data privacy for storing the data while performing transcript assembly. In this paper, we have developed a protocol named SecDATA which performs de novo transcript assembly ensuring data security. It consists of two modules. The first module deals with a framework for secured access and storage of data. The novelty of the first module lies in the employment of distributed ledger technology for data storage that ensures the privacy of the data. The second module deals with the development of an optimized graph-based method for de novo transcript assembly. We have compared our results with the state-of-art method de Bruijn graph and the popular pipeline Trinity, for transcript reconstruction, and our protocol outperforms them.

bioinformatics↗