Background: Metagenomic sequencing is becoming a powerful technology for exploring micro-ogranisms from various environments, such as human body, without isolation and cultivation. Accurately identifying genes from metagenomic fragments is one of the most fundamental issues. Results: In this article, we present a novel gene prediction method named MetaGUN for metagenomic fragments based on a machine learning approach of SVM. It implements in a three-stage strategy to predict genes. Firstly, it classifies input fragments into phylogenetic groups by a k-mer based sequence binning method. Then, proteincoding sequences are identified for each group independently with SVM classifiers that integrate entropy density profiles (EDP) of codon usage, translation initiation site (TIS) scores and open reading frame (ORF) length as input patterns. Finally, the TISs are adjusted by employing a modified version of MetaTISA. To identify protein-coding sequences, MetaGun builds the universal module and the novel module. The former is based on a set of representative species, while the latter is designed to find potential functionary DNA sequences with conserved domains.

BackgroundMetagenomic sequencing is becoming a powerful technology for exploring micro-ogranisms from various environments, such as human body, without isolation and cultivation. Accurately identifying genes from metagenomic fragments is one of the most fundamental issues.ResultsIn this article, we present a novel gene prediction method named MetaGUN for metagenomic fragments based on a machine learning approach of SVM. It implements in a three-stage strategy to predict genes. Firstly, it classifies input fragments into phylogenetic groups by a k-mer based sequence binning method. Then, protein-coding sequences are identified for each group independently with SVM classifiers that integrate entropy density profiles (EDP) of codon usage, translation initiation site (TIS) scores and open reading frame (ORF) length as input patterns. Finally, the TISs are adjusted by employing a modified version of MetaTISA. To identify protein-coding sequences, MetaGun builds the universal module and the novel module. The former is based on a set of representative species, while the latter is designed to find potential functionary DNA sequences with conserved domains.ConclusionsComparisons on artificial shotgun fragments with multiple current metagenomic gene finders show that MetaGUN predicts better results on both 3' and 5' ends of genes with fragments of various lengths. Especially, it makes the most reliable predictions among these methods. As an application, MetaGUN was used to predict genes for two samples of human gut microbiome. It identifies thousands of additional genes with significant evidences. Further analysis indicates that MetaGUN tends to predict more potential novel genes than other current metagenomic gene finders.

Background: Epileptic encephalopathies (EEs) are a pediatric entity with highly phenotypic and genetic heterogeneity. Both single nucleotide variants (SNVs)/Indels and copy number variations (CNVs) could be the causes. Whole exome sequencing (WES) is widely applied to detect SNVs/Indels, but the bioinformatics approach for detecting CNVs is still limited and weak. In the current study, the possibility of profiling both disease-causing SNVs/Indels and CNVs in a single test based on WES in EEs was evaluated.Methods: The infants diagnosed with EEs were enrolled from a single pediatric epilepsy center between January 2018 and February 2020. Demographic and clinical data were collected. In WES data, the pathogenic SNVs were identified through an in-house pipeline, and pathogenic CNVs were identified by CNVkit. The diagnostic rate was evaluated, and the molecular findings were characterized.Results: A total of 73 infants were included; 36 (49.32%) of them were males. The median age was 7 months. Thirty-two (43.84%) infants had been diagnosed with epilepsy syndrome. The most common type of syndrome was West syndrome (22/73, 30.1%), followed by Dravet syndrome (20/77, 27.4%). Fifty-four (73.97%) had intellectual development delay. The genetic cause of EEs, pathogenic or likely pathogenic variants, were successfully discovered in 46.6% (34/73) of the infants, and 29 (39.7%) infants carried SNVs/Indels, while 5 (6.8%) carried CNVs. The majority of the disease-causing variants were inherited in de novo pattern (25, 71.4%). In addition to showing that the variants in the ion channel encoding genes accounted for the main etiology, we discovered and confirmed two new disease-causing genes, CACNA1E and WDR26. Five discovered CNVs were deletions of 2q24.3, 1p36, 15q11-q13, 16p11.2, and 17p13.3, and all were confirmed by array comparative genomic hybridization.Conclusion: The application of both SNVs/Indels and CNVs detection in a single test based on WES yielded a high diagnosis rate in EEs. WES may serve as a first-tier test with cost-effective benefit in EEs.

Background: Short-chain enoyl-CoA hydratase deficiency (ECHS1D), also known as ECHS1 deficiency, is a rare inborn metabolic disorder with clinical presentations characterized by Leigh syndrome (LS). Thirty-four different pathogenic mutations have been identified from over 40 patients to date. Case presentation: Here, we report five Chinese patients with clinical syndromes typified as LS. Despite different initial symptoms, all patients presented developmental regression, dystonia, common radiological features such as symmetrical bilateral brain abnormalities, and similar metabolic results such as elevated plasma lactate and 2,3dihydroxy-2-methylbutyrate. Utilizing whole-exome sequencing (WES), we identified eight distinct variants in ECHS1, with six novel variants, and the remaining two variants have been previously reported. Interestingly, one of the six novel variants, c.463G > A (p.Gly155Ser), was detected in three patients from unrelated families, suggesting a potential founder effect already described for a few mutations in LS. Incorporating both genetic analysis and medical results, including magnetic resonance imaging (MRI), electroencephalography (EEG), and biochemical testing, our study enriched the mutation spectrum of the ECHS1 gene and confirmed the phenotypic presentations of LS. Conclusions: The severity of ECHS1 deficiency seems to vary. It was affected by both genetics and external environmental factors that lead to increased metabolism. Our study enriched the mutation spectrum of the ECHS1 gene, confirmed the phenotypic presentations, and highlighted the importance of the valine catabolic pathway in Leigh syndrome. Further studies are required to examine the potential founder mutation c.463G > A (p.Gly155Ser) and the role of ECHS1 in relevant pathways.

Summary We present the first tool of gene prediction, PlasGUN, for plasmid metagenomic short-read data. The tool, developed based on deep learning algorithm of multiple input Convolutional Neural Network, demonstrates much better performance when tested on a benchmark dataset of artificial short reads and presents more reliable results for real plasmid metagenomic data than traditional gene prediction tools designed primarily for chromosome-derived short reads. Availability and implementation The PlasGUN software is available at http://cqb.pku.edu.cn/ZhuLab/PlasGUN/ or https://github.com/zhenchengfang/PlasGUN/. Supplementary information Supplementary data are available at Bioinformatics online.