HELLO: improved neural network architectures and methodologies for small variant calling

• Published in: BMC bioinformatics 2021-08, Vol.22 (1), p.404-31, Article 404
• Main Authors: Ramachandran, Anand, Lumetta, Steven S, Klee, Eric W, Chen, Deming
• Format: Article
• Language: English
• Subjects: Accuracy; Algorithms; Analysis; Artificial neural networks; Computer architecture; Customization; Datasets; Deep learning; Deep neural networks; ESP; Experiments; Extrasensory perception; Gene loci; Genomes; Haplotypes; High-Throughput Nucleotide Sequencing; Humans; Hybrid variant calling; Illumina; INDEL Mutation; Machine learning; Methodology; Network architecture; Neural networks; Neural Networks, Computer; Object recognition; PacBio; Structure; Variant calling
• ISSN: 1471-2105; 1471-2105
• DOI: 10.1186/s12859-021-04311-4

Modern Next Generation- and Third Generation- Sequencing methods such as Illumina and PacBio Circular Consensus Sequencing platforms provide accurate sequencing data. Parallel developments in Deep Learning have enabled the application of Deep Neural Networks to variant calling, surpassing the accuracy of classical approaches in many settings. DeepVariant, arguably the most popular among such methods, transforms the problem of variant calling into one of image recognition where a Deep Neural Network analyzes sequencing data that is formatted as images, achieving high accuracy. In this paper, we explore an alternative approach to designing Deep Neural Networks for variant calling, where we use meticulously designed Deep Neural Network architectures and customized variant inference functions that account for the underlying nature of sequencing data instead of converting the problem to one of image recognition. Results from 27 whole-genome variant calling experiments spanning Illumina, PacBio and hybrid Illumina-PacBio settings suggest that our method allows vastly smaller Deep Neural Networks to outperform the Inception-v3 architecture used in DeepVariant for indel and substitution-type variant calls. For example, our method reduces the number of indel call errors by up to 18%, 55% and 65% for Illumina, PacBio and hybrid Illumina-PacBio variant calling respectively, compared to a similarly trained DeepVariant pipeline. In these cases, our models are between 7 and 14 times smaller. We believe that the improved accuracy and problem-specific customization of our models will enable more accurate pipelines and further method development in the field. HELLO is available at https://github.com/anands-repo/hello.