Metagenome shotgun sequence data offer the potential for large scale biogeographic analysis of microbial species. In this project we developed a method for detecting 33 common subtypes of the pathogenic bacterium Staphylococcus aureus. We used a binomial mixture model implemented in the binstrain software and the coverage counts at > 100,000 known S. aureus SNP (single nucleotide polymorphism) sites derived from prior comparative genomic analysis to estimate the proportion of each subtype in metagenome samples. Using this pipeline we were able to obtain > 87% sensitivity and > 94% specificity when testing on low genome coverage samples of diverse S. aureus strains (0.025X). We found that 321 and 149 metagenome samples from the Human Microbiome Project and metaSUB analysis of the New York City subway, respectively, contained S. aureus at genome coverage > 0.025. In both projects, CC8 and CC30 were the most common S. aureus subtypes encountered. We found evidence that the subtype composition at different body sites of the same individual were more similar than random sampling and more limited evidence that certain body sites were enriched for particular subtypes. One surprising finding was the apparent high frequency of CC398, a lineage associated with livestock, in samples from the tongue dorsum. Epidemiologic analysis of the HMP subject population suggested that high BMI (body mass index) and health insurance are risk factors for S. aureus but there was limited power to find factors linked to carriage of even the most common subtype. In the NYC subway data, we found a small signal of geographic distance affecting subtype clustering but other unknown factors influence taxonomic distribution of the species around the city. We argue that pathogen detection in metagenome samples requires the use of subtypes based on whole species population genomic analysis rather than using ad hoc collections of reference strains.

Summary: Advances in sequencing capacity have lead to the generation of unprecedented amounts of genomic data. The processing of this data frequently leads to I/O bottlenecks, e.g. when analyzing a small genomic region across a large number of samples. The largest I/O burden is, however, often not imposed by the amount of data needed for the analysis but rather by index files that help retrieving this data. We have developed chopBAI, a program that can chop a BAM index (BAI) file into small pieces. The program outputs a list of BAI files each indexing a specified genomic interval. The output files are much smaller in size but maintain compatibility with existing software tools. We show how preprocessing BAI files with chopBAI can lead to a reduction of I/O by more than 95 % during the analysis of 10 Kbp genomic regions, eventually enabling the joint analysis of more than 10,000 individuals. Availability and Implementation: The software is implemented in C++, GPL licensed and available at http://github.com/DecodeGenetics/chopBAI Contact: birte.kehr@decode.is

Whole genome sequencing on next-generation instruments provides an unbiased way to identify the organisms present in complex metagenomic samples. However, the time-to-result can be protracted because of fixed-time sequencing runs and cumbersome bioinformatics workflows. This limits the utility of the approach in settings where rapid species identification is crucial, such as in the quality control of food-chain components, or in during an outbreak of an infectious disease. Here we present What′s in my Pot? (WIMP), a laboratory and analysis workflow in which, starting with an unprocessed sample, sequence data is generated and bacteria, viruses and fungi present in the sample are classified to subspecies and strain level in a quantitative manner, without prior knowledge of the sample composition, in approximately 3.5 hours. This workflow relies on the combination of Oxford Nanopore Technologies′ MinION ™ sensing device with a real-time species identification bioinformatics application.

Shared genetic architecture between phenotypes may be driven by a common genetic basis (pleiotropy) or a subset of genetically similar individuals (heterogeneity). We developed BUHMBOX, a well-powered statistical method to distinguish pleiotropy from heterogeneity using genotype data. We observed a shared genetic basis between 11 of 17 tested autoimmune diseases and type I diabetes (T1D, p<10 12) and 11 of 17 tested autoimmune diseases and rheumatoid arthritis (RA, p<10-7). This sharing could not be explained by heterogeneity (corrected pBUHMBOX>0.2 using 6,670 T1D cases and 7,279 RA cases), suggesting that shared genetic features in autoimmunity are due to pleiotropy. We observed a shared genetic basis between seronegative and seropostive RA (p<10-22), explained by heterogeneity (pBUHMBOX=0.008 in 2,406 seronegative RA cases). Consistent with previous observations, we observed genetic sharing between major depressive disorder (MDD) and schizophrenia (p<10 9). This sharing is not explained by heterogeneity (pBUHMBOX=0.28 in 9,238 MDD cases).

While eukaryotic noncoding RNAs have recently received intense scrutiny, it is becoming clear that bacterial transcription is at least as pervasive. Bacterial small RNAs and antisense RNAs (sRNAs) are often assumed to be noncoding, due to their lack of long open reading frames (ORFs). However, there are numerous examples of sRNAs encoding for small proteins, whether or not they also have a regulatory role at the RNA level. Here, we apply flexible machine learning techniques based on sequence features and comparative genomics to quantify the prevalence of sRNA ORFs under natural selection to maintain protein-coding function in phylogenetically diverse bacteria. A majority of annotated sRNAs have at least one ORF between 10 and 50 amino acids long, and we conservatively predict that 188 ± 25.5 unannotated sRNA ORFs are under selection to maintain coding, an average of 13 per species considered here. This implies that overall at least 7.5 ± 0.3% of sRNAs have a coding ORF, and in some species at least 20% do. 84 ± 9.8 of these novel coding ORFs have some antisense overlap to annotated ORFs. As experimental validation, many of our predictions are translated according to ribosome profiling data and are identified via mass spectrometry shotgun proteomics. B. subtilis sRNAs with coding ORFs are enriched for high expression in biofilms and confluent growth, and two S. pneumoniae sRNAs with coding ORFs are involved in virulence. sRNA coding ORFs are enriched for transmembrane domains and many are novel components of type I toxin/antitoxin systems. Our predictions for sRNA coding ORFs, including novel type I toxins, are freely available in a user-friendly format at http://disco-bac.web.pasteur.fr.

Whereas polyploidy is common and an important evolutionary factor in most land plant lineages it is a real rarity in gymnosperms. Coast redwood (Sequoia sempervirens) is the only hexaploid conifer and one of just two naturally polyploid conifer species. Numerous hypotheses about the mechanism of polyploidy in Sequoia and parental genome donors have been proffered over the years, primarily based on morphological and cytological data, but it remains unclear how Sequoia became polyploid and why this lineage overcame an apparent gymnosperm barrier to whole-genome duplication (WGD). We sequenced transcriptomes and used phylogenetic inference, Bayesian concordance analysis, and paralog age distributions to resolve relationships among gene copies in hexaploid coast redwood and its close relatives. Our data show that hexaploidy in the coast redwood lineage is best explained by autopolyploidy or, if there was allopolyploidy, this was restricted to within the Californian redwood clade. We found that duplicate genes have more similar sequences than would be expected given evidence from fossil guard cell size which suggest that polyploidy dates to the Eocene. Conflict between molecular and fossil estimates of WGD can be explained if diploidization occurred very slowly following whole genome duplication. We extrapolate from this to suggest that the rarity of polyploidy in conifers may be due to slow rates of diploidization in this clade.

Recent emergence of nanopore sequencing technology set a challenge for the established assembly methods not optimized for the combination of read lengths and high error rates of nanopore reads. In this work we assessed how existing de novo assembly methods perform on these reads. We benchmarked three non-hybrid (in terms of both error correction and scaffolding) assembly pipelines as well as two hybrid assemblers which use third generation sequencing data to scaffold Illumina assemblies. Tests were performed on several publicly available MinION and Illumina datasets of E. coli K-12, using several sequencing coverages of nanopore data (20x, 30x, 40x and 50x). We attempted to assess the quality of assembly at each of these coverages, to estimate the requirements for closed bacterial genome assembly. Results show that hybrid methods are highly dependent on the quality of NGS data, but much less on the quality and coverage of nanopore data and perform relatively well on lower nanopore coverages. Furthermore, when coverage is above 40x, all non-hybrid methods correctly assemble the E. coli genome, even a non-hybrid method tailored for Pacific Bioscience reads. While it requires higher coverage compared to a method designed particularly for nanopore reads, its running time is significantly lower.