iPRIns: A Tool with the Improved Precision and Recall for Insertion Detection in the Human Genome

An insertion is a specific type of the structural variations. The identification of insertions in a human genome is essential for the study of diseases or their functional effects. There are many tools available for identifying the insertion type with different methods and strategies. However, most of them could not deliver both good recall and precision, especially for the real datasets sequenced with the paired-end short reads. In this paper, we propose iPRIns, a new computational method for detecting insertions aiming to improve both precision and recall. The proposed method with discovering and filtering processes outperformed all other three tools for 5 out of 10 real datasets, the variations of NA12878, for both precision and recall. iPRIns is released under the open-source GPLv3 license. The source code and documentation are available at https://github.com/cucpbioinfo/iPRIns.