My Work

TIR-Learner v3

The Ou Lab, The Ohio State University @ Columbus, OH

Download Source Code

Abstract

TIR-Learner is an ensemble pipeline for Terminal Inverted Repeat (TIR) transposable elements annotation, which combines homology-based detection, de novo tools and convolutional neural network to classify candidate sequences into five major TIR superfamilies. The old TIR suffers from slow execution on large genomes due to intense I/O operations and less efficient algorithms, it also lacks maintainability due to legacy dependency issues. Undergone a completely re-design and rewrite over the past year, focusing on enhancing efficiency, improving compatibility and ensuring code quality, both the efficiency and the speed of the program has been elevated by 32%.

Benchmark Result

Benchmark result of execution time on 100MiB test file
Benchmark result of execution time on 300MiB test file
Benchmark result of CPU utilization

Sponsorship

The logo of The Ohio State University

OSU undergraduate research access innovation seed grant

 The logo of Ohio Supercomputer Center

Ohio Supercomputer Center

Quick Links