BACKGROUND: Shotgun metagenomes are often assembled prior to annotation of genes which biases the functional capacity of a community towards its most abundant members. For an unbiased assessment of community function, short reads need to be mapped directly to a gene or protein database. The ability to detect genes in short read sequences is dependent on pre- and post-sequencing decisions. The objective of the current study was to determine how library size selection, read length and format, protein database, e-value threshold, and sequencing depth impact gene-centric analysis of human fecal microbiomes when using DIAMOND, an alignment tool that is up to 20,000 times faster than BLASTX. RESULTS: Using metagenomes simulated from a database of experimentally verified protein sequences, we find that read length, e-value threshold, and the choice of protein database dramatically impact detection of a known target, with best performance achieved with longer reads, stricter e-value thresholds, and a custom database. Using publicly available metagenomes, we evaluated library size selection, paired end read strategy, and sequencing depth. Longer read lengths were acheivable by merging paired ends when the sequencing library was size-selected to enable overlaps. When paired ends could not be merged, a congruent strategy in which both ends are independently mapped was acceptable. Sequencing depths of 5 million merged reads minimized the error of abundance estimates of specific target genes, including an antimicrobial resistance gene. CONCLUSIONS: Shotgun metagenomes of DNA extracted from human fecal samples sequenced using the Illumina platform should be size-selected to enable merging of paired end reads and should be sequenced in the PE150 format with a minimum sequencing depth of 5 million merge-able reads to enable detection of specific target genes. Expecting the merged reads to be 180-250 bp in length, the appropriate e-value threshold for DIAMOND would then need to be more strict than the default. Accurate and interpretable results for specific hypotheses will be best obtained using small databases customized for the research question.

BACKGROUND: Complex microbial communities are an area of growing interest in biology. Metatranscriptomics allows researchers to quantify microbial gene expression in an environmental sample via high-throughput sequencing. Metatranscriptomic experiments are computationally intensive because the experiments generate a large volume of sequence data and each sequence must be compared with reference sequences from thousands of organisms. RESULTS: SAMSA2 is an upgrade to the original Simple Annotation of Metatranscriptomes by Sequence Analysis (SAMSA) pipeline that has been redesigned for standalone use on a supercomputing cluster. SAMSA2 is faster due to the use of the DIAMOND aligner, and more flexible and reproducible because it uses local databases. SAMSA2 is available with detailed documentation, and example input and output files along with examples of master scripts for full pipeline execution. CONCLUSIONS: SAMSA2 is a rapid and efficient metatranscriptome pipeline for analyzing large RNA-seq datasets in a supercomputing cluster environment. SAMSA2 provides simplified output that can be examined directly or used for further analyses, and its reference databases may be upgraded, altered or customized to fit the needs of any experiment.

BACKGROUND: Although metatranscriptomics-the study of diverse microbial population activity based on RNA-seq data-is rapidly growing in popularity, there are limited options for biologists to analyze this type of data. Current approaches for processing metatranscriptomes rely on restricted databases and a dedicated computing cluster, or metagenome-based approaches that have not been fully evaluated for processing metatranscriptomic datasets. We created a new bioinformatics pipeline, designed specifically for metatranscriptome dataset analysis, which runs in conjunction with Metagenome-RAST (MG-RAST) servers. Designed for use by researchers with relatively little bioinformatics experience, SAMSA offers a breakdown of metatranscriptome transcription activity levels by organism or transcript function, and is fully open source. We used this new tool to evaluate best practices for sequencing stool metatranscriptomes. RESULTS: Working with the MG-RAST annotation server, we constructed the Simple Annotation of Metatranscriptomes by Sequence Analysis (SAMSA) software package, a complete pipeline for the analysis of gut microbiome data. SAMSA can summarize and evaluate raw annotation results, identifying abundant species and significant functional differences between metatranscriptomes. Using pilot data and simulated subsets, we determined experimental requirements for fecal gut metatranscriptomes. Sequences need to be either long reads (longer than 100 bp) or joined paired-end reads. Each sample needs 40-50 million raw sequences, which can be expected to yield the 5-10 million annotated reads necessary for accurate abundance measures. We also demonstrated that ribosomal RNA depletion does not equally deplete ribosomes from all species within a sample, and remaining rRNA sequences should be discarded. Using publicly available metatranscriptome data in which rRNA was not depleted, we were able to demonstrate that overall organism transcriptional activity can be measured using mRNA counts. We were also able to detect significant differences between control and experimental groups in both organism transcriptional activity and specific cellular functions. CONCLUSIONS: By making this new pipeline publicly available, we have created a powerful new tool for metatranscriptomics research, offering a new method for greater insight into the activity of diverse microbial communities. We further recommend that stool metatranscriptomes be ribodepleted and sequenced in a 100 bp paired end format with a minimum of 40 million reads per sample.

UNLABELLED: Hidden Markov models (HMMs) are probabilistic models that are well-suited to solve many different classification problems in computation biology. StochHMM provides a command-line program and C++ library that can implement a traditional HMM from a simple text file. StochHMM provides researchers the flexibility to create higher-order emissions, integrate additional data sources and/or user-defined functions into multiple points within the HMM framework. Additional features include user-defined alphabets, ability to handle ambiguous characters in an emission-dependent manner, user-defined weighting of state paths and ability to tie transition probabilities to sequence. AVAILABILITY AND IMPLEMENTATION: StochHMM is implemented in C++ and is available under the MIT License. Software, source code, documentation and examples can be found at http://github.com/KorfLab/StochHMM.

BACKGROUND: In previous studies, gene neighborhoods-spatial clusters of co-expressed genes in the genome-have been defined using arbitrary rules such as requiring adjacency, a minimum number of genes, a fixed window size, or a minimum expression level. In the current study, we developed a Gene Neighborhood Scoring Tool (G-NEST) which combines genomic location, gene expression, and evolutionary sequence conservation data to score putative gene neighborhoods across all possible window sizes simultaneously. RESULTS: Using G-NEST on atlases of mouse and human tissue expression data, we found that large neighborhoods of ten or more genes are extremely rare in mammalian genomes. When they do occur, neighborhoods are typically composed of families of related genes. Both the highest scoring and the largest neighborhoods in mammalian genomes are formed by tandem gene duplication. Mammalian gene neighborhoods contain highly and variably expressed genes. Co-localized noisy gene pairs exhibit lower evolutionary conservation of their adjacent genome locations, suggesting that their shared transcriptional background may be disadvantageous. Genes that are essential to mammalian survival and reproduction are less likely to occur in neighborhoods, although neighborhoods are enriched with genes that function in mitosis. We also found that gene orientation and protein-protein interactions are partially responsible for maintenance of gene neighborhoods. CONCLUSIONS: Our experiments using G-NEST confirm that tandem gene duplication is the primary driver of non-random gene order in mammalian genomes. Non-essentiality, co-functionality, gene orientation, and protein-protein interactions are additional forces that maintain gene neighborhoods, especially those formed by tandem duplicates. We expect G-NEST to be useful for other applications such as the identification of core regulatory modules, common transcriptional backgrounds, and chromatin domains. The software is available at http://docpollard.org/software.html.

MOTIVATION: The numbers of finished and ongoing genome projects are increasing at a rapid rate, and providing the catalog of genes for these new genomes is a key challenge. Obtaining a set of well-characterized genes is a basic requirement in the initial steps of any genome annotation process. An accurate set of genes is needed in order to learn about species-specific properties, to train gene-finding programs, and to validate automatic predictions. Unfortunately, many new genome projects lack comprehensive experimental data to derive a reliable initial set of genes. RESULTS: In this study, we report a computational method, CEGMA (Core Eukaryotic Genes Mapping Approach), for building a highly reliable set of gene annotations in the absence of experimental data. We define a set of conserved protein families that occur in a wide range of eukaryotes, and present a mapping procedure that accurately identifies their exon-intron structures in a novel genomic sequence. CEGMA includes the use of profile-hidden Markov models to ensure the reliability of the gene structures. Our procedure allows one to build an initial set of reliable gene annotations in potentially any eukaryotic genome, even those in draft stages. AVAILABILITY: Software and data sets are available online at http://korflab.ucdavis.edu/Datasets.

BACKGROUND: Computational gene prediction continues to be an important problem, especially for genomes with little experimental data. RESULTS: I introduce the SNAP gene finder which has been designed to be easily adaptable to a variety of genomes. In novel genomes without an appropriate gene finder, I demonstrate that employing a foreign gene finder can produce highly inaccurate results, and that the most compatible parameters may not come from the nearest phylogenetic neighbor. I find that foreign gene finders are more usefully employed to bootstrap parameter estimation and that the resulting parameters can be highly accurate. CONCLUSION: Since gene prediction is sensitive to species-specific parameters, every genome needs a dedicated gene finder.

MOTIVATION: The translating BLAST algorithms are powerful tools for finding protein-coding genes because they identify amino acid similarities in nucleotide sequences. Unfortunately, these kinds of searches are computationally intensive and often represent bottlenecks in sequence analysis pipelines. Tuning parameters for speed can make the searches much faster, but one risks losing low-scoring alignments. However, high scoring alignments are relatively resistant to such changes in parameters, and this fact makes it possible to use a serial strategy where a fast, insensitive search is used to pre-screen a database for similar sequences, and a slow, sensitive search is used to produce the sequence alignments. RESULTS: Serial BLAST searches improve both the speed and sensitivity.

Abstract Summary: Identifying and masking repetitive elements is usually the first step when analyzing vertebrate genomic sequence. Current repeat identification software is sensitive but slow, creating a costly bottleneck in large-scale analyses. We have developed MaskerAid , a software enhancement to RepeatMasker that increased the speed of masking more than 30-fold at the most sensitive setting. Availability: On request from the authors (see http://sapiens.wustl.edu/MaskerAid). Contact: [email protected] These authors contributed equally to this work. * To whom correspondence should be addressed. ** Current address: Incyte Genomics, 4633 World Parkway Cir., St. Louis, MO 63134 USA, [email protected]

UNLABELLED: We have developed a program, MPBLAST, that increases the throughput of batch BLASTN searches by multiplexing (concatenating) query sequences and thereby reducing the number of actual database searches performed. Throughput was observed to increase in reciprocal proportion to the component sequence length. For sequencing read-sized queries of 500 bp, an order of magnitude speed-up was seen. AVAILABILITY: Free (see http://blast.wustl.edu) CONTACT: [ikorf, gish]@watson.wustl.edu