Le, V. A. | Search bioRxiv

bioRxiv2024

Particular sequence characteristics induce bias in the detection of polymorphic transposable element insertions

Transposable elements (TEs) have an important role in genome evolution but are challenging for bioinformatics detection due to their repetitive nature and ability to move and replicate within genomes. New sequencing technologies now enable the characterization of nucleotide and structural variations within species. Among them, TE polymorphism is critical to identify as it may influence species adaptation or trigger diseases. Despite the development of numerous bioinformatic programs, identifying the most effective tool is challenging due to non-overlapping results and varying efficiency across studies. Benchmarking efforts have highlighted some of the limitations of these tools, often evaluated on either real or simulated data. However, real data may be incomplete or contain unannotated TEs, while simulated data may not accurately reflect real genomes. This study introduces a simulation method generating data based on real genomes to control all genomic parameters. Evaluating several TE polymorphic detection tools using data from Drosophila melanogaster and Arabidopsis thaliana, our study investigates factors like copy size, sequence divergence, and GC content that influence detection efficiency. Our results indicate that only a few programs perform satisfactorily and that all are sensitive to TE and genomic characteristics that may differ according to the species considered. Using Bos taurus population data as a case study to identify polymorphic LTR-retrotransposon insertions, we found low-frequency insertions particularly challenging to detect due to a high number of false positives. Increased sequencing coverage improved sensitivity but reduced precision. Our work underscores the importance of selecting appropriate tools and thresholds according to the specific research questions.

bioinformatics↗

bioRxiv2021

Markerless mouse tracking for social experiments

Automated behavior quantification requires accurate tracking of animals. Simultaneous tracking of multiple animals, particularly those lacking visual identifiers, is particularly challenging. Problems of mistaken identities and lost information on key anatomical features are common in existing methods. Here we propose a markerless video-based tool to simultaneously track two socially interacting mice of the same appearance. It incorporates conventional handcrafted tracking and deep learning based techniques, which are trained on a small number of labeled images from a very basic, uncluttered experimental setup. The output consists of body masks and coordinates of the snout and tail-base for each mouse. The method was tested on a series of cross-setup videos recorded under commonly used experimental conditions including bedding in the cage and fiberoptic or headstage implants on the mice. Results obtained without any human intervention showed the effectiveness of the proposed approach, evidenced by a near elimination of identities switches and a 10% improvement in tracking accuracy over a pure deep-learning-based keypoint tracking approach trained on the same data. Finally, we demonstrated an application of this approach in studies of social behaviour of mice, by using it to quantify and compare interactions between pairs of mice in which some are anosmic, i.e. unable to smell. Our results indicated loss of olfaction impaired typical snout-directed social recognition behaviors of mice, while non-snout-directed social behaviours were enhanced. Together, these results suggest that the hybrid approach could be valuable for studying group behaviors in rodents, such as social interactions.

bioengineering↗