Quantitative analysis of transcription start site selection reveals control by DNA sequence, RNA polymerase II activity and NTP levels.

Transcription start site (TSS) selection is a key step in gene expression and occurs at many promoter positions over a wide range of efficiencies. Here we develop a massively parallel reporter assay to quantitatively dissect contributions of promoter sequence, nucleoside triphosphate substrate levels and RNA polymerase II (Pol II) activity to TSS selection by 'promoter scanning' in Saccharomyces cerevisiae (Pol II MAssively Systematic Transcript End Readout, 'Pol II MASTER'). Using Pol II MASTER, we measure the efficiency of Pol II initiation at 1,000,000 individual TSS sequences in a defined promoter context. Pol II MASTER confirms proposed critical qualities of S. cerevisiae TSS -8, -1 and +1 positions, quantitatively, in a controlled promoter context. Pol II MASTER extends quantitative analysis to surrounding sequences and determines that they tune initiation over a wide range of efficiencies. These results enabled the development of a predictive model for initiation efficiency based on sequence. We show that genetic perturbation of Pol II catalytic activity alters initiation efficiency mostly independently of TSS sequence, but selectively modulates preference for the initiating nucleotide. Intriguingly, we find that Pol II initiation efficiency is directly sensitive to guanosine-5'-triphosphate levels at the first five transcript positions and to cytosine-5'-triphosphate and uridine-5'-triphosphate levels at the second position genome wide. These results suggest individual nucleoside triphosphate levels can have transcript-specific effects on initiation, representing a cryptic layer of potential regulation at the level of Pol II biochemical properties. The results establish Pol II MASTER as a method for quantitative dissection of transcription initiation in eukaryotes.

Higher-order epistasis within Pol II trigger loop haplotypes.

RNA polymerase II (Pol II) has a highly conserved domain, the trigger loop (TL), that controls transcription fidelity and speed. We previously probed pairwise genetic interactions between residues within and surrounding the TL and identified widespread incompatibility between TLs of different species when placed in the Saccharomyces cerevisiae Pol II context, indicating epistasis between the TL and its surrounding context. We sought to understand the nature of this incompatibility and probe higher order epistasis internal to the TL. We have employed deep mutational scanning with selected natural TL variants ("haplotypes"), and all possible intermediate substitution combinations between them and the yeast Pol II TL. We identified both positive and negative higher-order residue interactions within example TL haplotypes. Intricate higher-order epistasis formed by TL residues was sometimes only apparent from analysis of intermediate genotypes, emphasizing complexity of epistatic interactions. Furthermore, we distinguished TL substitutions with distinct classes of epistatic patterns, suggesting specific TL residues that potentially influence TL evolution. Our examples of complex residue interactions suggest possible pathways for epistasis to facilitate Pol II evolution.

Widespread epistasis shapes RNA Polymerase II active site function and evolution.

Multi-subunit RNA Polymerases (msRNAPs) are responsible for transcription in all kingdoms of life. At the heart of these msRNAPs is an ultra-conserved active site domain, the trigger loop (TL), coordinating transcription speed and fidelity by critical conformational changes impacting multiple steps in substrate selection, catalysis, and translocation. Previous studies have observed several different types of genetic interactions between eukaryotic RNA polymerase II (Pol II) TL residues, suggesting that the TL's function is shaped by functional interactions of residues within and around the TL. The extent of these interaction networks and how they control msRNAP function and evolution remain to be determined. Here we have dissected the Pol II TL interaction landscape by deep mutational scanning in Saccharomyces cerevisiae Pol II. Through analysis of over 15000 alleles, representing all single mutants, a rationally designed subset of double mutants, and evolutionarily observed TL haplotypes, we identify interaction networks controlling TL function. Substituting residues creates allele-specific networks and propagates epistatic effects across the Pol II active site. Furthermore, the interaction landscape further distinguishes alleles with similar growth phenotypes, suggesting increased resolution over the previously reported single mutant phenotypic landscape. Finally, co-evolutionary analyses reveal groups of co-evolving residues across Pol II converge onto the active site, where evolutionary constraints interface with pervasive epistasis. Our studies provide a powerful system to understand the plasticity of RNA polymerase mechanism and evolution, and provide the first example of pervasive epistatic landscape in a highly conserved and constrained domain within an essential enzyme.

Identification and Characterization of Small RNA Markers of Age in the Blow Fly Cochliomyia macellaria (Fabricius) (Diptera: Calliphoridae).

Blow fly development is important in decomposition ecology, agriculture, and forensics. Much of the impact of these species is from immature samples, thus knowledge of their development is important to enhance or ameliorate their effects. One application of this information is the estimation of immature insect age to provide temporal information for death investigations. While traditional markers of age such as stage and size are generally accurate, they lack precision in later developmental stages. We used miRNA sequencing to measure miRNA expression, throughout development, of the secondary screwworm, Cochliomyia macellaria (Fabricius) (Diptera: Calliphoridae) and identified 217 miRNAs present across the samples. Ten were identified to be significantly differentially expressed in larval samples and seventeen were found to be significantly differentially expressed in intrapuparial samples. Twenty-eight miRNAs were identified to be differentially expressed between sexes. Expression patterns of two miRNAs, miR-92b and bantam, were qPCR-validated in intrapuparial samples; these and likely food-derived miRNAs appear to be stable markers of age in C. macellaria. Our results support the use of miRNAs for developmental markers of age and suggest further investigations across species and under a range of abiotic and biotic conditions.

Genome-wide identification of genes enabling accurate prediction of hybrid performance from parents across environments and populations for gene-based breeding in maize.

Accurate prediction of hybrid offspring complex trait phenotype from parents is paramount to enhanced plant breeding, animal breeding, and human medicine. Here we report genome-wide identification of genes enabling accurate prediction of hybrid offspring complex traits from parents using maize grain yield as the target trait. We identified 181 ZmF1GY genes enabling prediction of maize (Zea mays L.) F1 hybrid grain yield from parents and tested their utility and efficiency for predicting F1 hybrid grain yields from parents using their expressions, genic SNPs, and number of favorable alleles (NFAs), respectively. The ZmF1GY genes predicted hybrid grain yields from parents at an accuracy of 0.86, presented by correlation coefficient between predicted and observed phenotypes, within an environment, 0.74 across environments, and 0.64 across populations, outperforming genomic prediction by 27-406%, 23%, and 40%, respectively. Furthermore, we identified nine of the ZmF1GY genes containing SNPs or InDels in parents that increased or decreased hybrid grain yields by 14-46%. When the NFAs of these nine ZmF1GY genes were used for hybrid grain yield prediction from parents, they predicted hybrid grain yields at an accuracy of 0.79, outperforming genomic prediction by 21% that was based on up to tens of thousands of genome-wide SNPs. These results demonstrate the feasibility of developing a gene toolkit for a species enabling gene-based breeding across environments and populations that is much more powerful and efficient than current breeding, thereby helping secure the world's food production. The methodology is applicable to all crops, livestock, and humans.

Hermetia illucens L. larvae-associated intestinal microbes reduce the transmission risk of zoonotic pathogens in pig manure.

Black soldier fly (BSF) larvae are considered a promising biological reactor to convert organic waste and reduce the impact of zoonotic pathogens on the environment. We analysed the effects of BSF larvae on Staphylococcus aureus and Salmonella spp. populations in pig manure (PM), which showed that BSF larvae can significantly reduce the counts of the associated S. aureus and Salmonella spp. Then, using a sterile BSF larval system, we validated the function of BSF larval intestinal microbiota in vivo to suppress pathogens, and lastly, we isolated eight bacterial strains from the BSF larval gut that inhibit S. aureus. Results indicated that functional microbes are essential for BSF larvae to antagonise S. aureus. Moreover, the analysis results of the relationship between the intestinal microbiota and S. aureus and Salmonella spp. showed that Myroides, Tissierella, Oblitimonas, Paenalcalignes, Terrisporobacter, Clostridium, Fastidiosipila, Pseudomonas, Ignatzschineria, Savagea, Moheibacter and Sphingobacterium were negatively correlated with S. aureus and Salmonella. Overall, these results suggested that the potential ability of BSF larvae to inhibit S. aureus and Salmonella spp. present in PM is accomplished primarily by gut-associated microorganisms.

Analysis of the genes controlling cotton fiber length reveals the molecular basis of plant breeding and the genetic potential of current cultivars for continued improvement.

Stagnated crop improvement has raised questions of whether and how current crop cultivars can be further improved. Genes are the core determinants of performance of all cultivars. Here, we report the molecular basis of plant breeding and address these questions by analyzing 226 GFL genes controlling and accurately predicting fiber length, an important breeding objective trait, in cotton (Gossypium sp.). We first identified the favorable allele and the number of favorable alleles (NFAs) of each GFL gene, calculated the total NFAs of the 226 GFL genes accumulated in 198 advanced breeding lines, and analyzed them against fiber lengths. Fiber lengths of the breeding lines were strongly correlated with the total NFAs of the GFL genes (r = 0.85, P < 0.0001), suggesting that accumulation of the favorable alleles of the genes controlling objective traits is the molecular basis of cotton breeding. Surprisingly, a breeding line with a fiber length of present cultivars having the longest fibers contained only about 51% of the total NFAs of the 226 GFL genes. The genetic potentials of current cultivars were then predicted using linear and non-linear models, respectively, revealing that a breeding line or cultivar with a fiber length of 33.8 mm could be further improved in fiber length by up to 118%. Finally, we showed that the genetic potential of such a breeding line can be realized through gene-based breeding. Therefore, these findings shed light on continued crop improvement in general and provide 740 genic biomarkers desirable for enhanced cotton fiber breeding.

Brucella activates the host RIDD pathway to subvert BLOS1-directed immune defense.

The phagocytosis and destruction of pathogens in lysosomes constitute central elements of innate immune defense. Here, we show that Brucella, the causative agent of brucellosis, the most prevalent bacterial zoonosis globally, subverts this immune defense pathway by activating regulated IRE1α-dependent decay (RIDD) of Bloc1s1 mRNA encoding BLOS1, a protein that promotes endosome-lysosome fusion. RIDD-deficient cells and mice harboring a RIDD-incompetent variant of IRE1α were resistant to infection. Inactivation of the Bloc1s1 gene impaired the ability to assemble BLOC-1-related complex (BORC), resulting in differential recruitment of BORC-related lysosome trafficking components, perinuclear trafficking of Brucella-containing vacuoles (BCVs), and enhanced susceptibility to infection. The RIDD-resistant Bloc1s1 variant maintains the integrity of BORC and a higher-level association of BORC-related components that promote centrifugal lysosome trafficking, resulting in enhanced BCV peripheral trafficking and lysosomal destruction, and resistance to infection. These findings demonstrate that host RIDD activity on BLOS1 regulates Brucella intracellular parasitism by disrupting BORC-directed lysosomal trafficking. Notably, coronavirus murine hepatitis virus also subverted the RIDD-BLOS1 axis to promote intracellular replication. Our work establishes BLOS1 as a novel immune defense factor whose activity is hijacked by diverse pathogens.

Accurate prediction of complex traits for individuals and offspring from parents using a simple, rapid, and efficient method for gene-based breeding in cotton and maize.

Accurate, simple, rapid, and inexpensive prediction of complex traits controlled by numerous genes is paramount to enhanced plant breeding, animal breeding, and human medicine. Here we report a novel method that enables accurate, simple, and rapid prediction of complex traits of individuals or offspring from parents based on the number of favorable alleles (NFAs) of the genes controlling the objective traits. The NFAs of 226 cotton fiber length (GFL) genes and nine maize hybrid grain yield related (ZmF1GY) genes were directly used to predict cotton fiber lengths of individual plants and maize grain yields of F1 hybrids from parents, respectively, using prediction model-based methods as controls. The NFAs of the 226 GFL genes predicted cotton fiber lengths at an accuracy of 0.85, as the model methods and outperforming genomic prediction by 82 % - 170 %. The NFAs of the nine ZmF1GY genes predicted grain yields of maize hybrids from parents at an accuracy of 0.80, outperforming genomic prediction by 67 %. Moreover, the prediction accuracies of these traits were consistent across years, environments, and eco-agricultural systems. Importantly, the accurate prediction of these traits directly using the NFAs of the genes allows breeding to be performed in greenhouse, phytotron, or off-season, without the need of the model training and validation steps essential and costly for model-based genomic or genic prediction. Therefore, this new method dramatically outperforms the current model-based genomic methods used for phenotype prediction and streamlines the process of breeding, thus promising to substantially enhance current plant and animal breeding.

Molecular and genetic dissection of the USDA rice mini-core collection using high-density SNP markers.

Molecular tools and knowledge of crop germplasm are vital for their effective utilization. In this study, we developed 40,866 high-quality and well distributed SNPs for a rice mini-core collection (RMC) developed by the United States Department of Agriculture (USDA). The high-quality SNPs clustered the USDA-RMC into five subpopulations (Ind, indica; Aus, aus; Afr, African rice; TeJ, temperate japonica; TrJ, tropical japonica) and one admixture (Adm). This classification was further confirmed by phylogenetic and principal component analyses. The rice ARO (aromatic) subpopulation of previous studies was re-assigned with Adm and the WD (wild-type) subpopulation was re-defined to the Afr subpopulation because most of its accessions are African cultivated rice. The Aus and Ind subpopulations had a substantially wider genetic variation than the TrJ and TeJ subpopulations. The genetic diversities were much larger between the Ind or Aus subpopulation and the TrJ or TeJ subpopulation than between the Afr subpopulation and the Ind, Aus, TrJ or TeJ subpopulation. Comparative agronomic trait analysis between the subpopulations also supported the genetic structure and variation of the RMC, and suggested the existence of extensive variation in the genes controlling agronomic traits among them. Furthermore, analysis of ancestral membership of the RMC accessions revealed that reproductive barrier or wide incompatibility existed between the Indica and Japonica groups, while gene flow occurred between them. These results provide high-quality SNPs and knowledge of genetic structure and diversity of the USDA-RMC necessary for enhanced rice research and breeding.

The genome of the stable fly, Stomoxys calcitrans, reveals potential mechanisms underlying reproduction, host interactions, and novel targets for pest control.

BACKGROUND: The stable fly, Stomoxys calcitrans, is a major blood-feeding pest of livestock that has near worldwide distribution, causing an annual cost of over $2 billion for control and product loss in the USA alone. Control of these flies has been limited to increased sanitary management practices and insecticide application for suppressing larval stages. Few genetic and molecular resources are available to help in developing novel methods for controlling stable flies. RESULTS: This study examines stable fly biology by utilizing a combination of high-quality genome sequencing and RNA-Seq analyses targeting multiple developmental stages and tissues. In conjunction, 1600 genes were manually curated to characterize genetic features related to stable fly reproduction, vector host interactions, host-microbe dynamics, and putative targets for control. Most notable was characterization of genes associated with reproduction and identification of expanded gene families with functional associations to vision, chemosensation, immunity, and metabolic detoxification pathways. CONCLUSIONS: The combined sequencing, assembly, and curation of the male stable fly genome followed by RNA-Seq and downstream analyses provide insights necessary to understand the biology of this important pest. These resources and new data will provide the groundwork for expanding the tools available to control stable fly infestations. The close relationship of Stomoxys to other blood-feeding (horn flies and Glossina) and non-blood-feeding flies (house flies, medflies, Drosophila) will facilitate understanding of the evolutionary processes associated with development of blood feeding among the Cyclorrhapha.

Accurate Prediction of a Quantitative Trait Using the Genes Controlling the Trait for Gene-Based Breeding in Cotton.

Accurate phenotype prediction of quantitative traits is paramount to enhanced plant research and breeding. Here, we report the accurate prediction of cotton fiber length, a typical quantitative trait, using 474 cotton (Gossypium ssp.) fiber length (GFL) genes and nine prediction models. When the SNPs/InDels contained in 226 of the GFL genes or the expressions of all 474 GFL genes was used for fiber length prediction, a prediction accuracy of r = 0.83 was obtained, approaching the maximally possible prediction accuracy of a quantitative trait. This has improved by 116%, the prediction accuracies of the fiber length thus far achieved for genomic selection using genome-wide random DNA markers. Moreover, analysis of the GFL genes identified 125 of the GFL genes that are key to accurate prediction of fiber length, with which a prediction accuracy similar to that of all 474 GFL genes was obtained. The fiber lengths of the plants predicted with expressions of the 125 key GFL genes were significantly correlated with those predicted with the SNPs/InDels of the above 226 SNP/InDel-containing GFL genes (r = 0.892, P = 0.000). The prediction accuracies of fiber length using both genic datasets were highly consistent across environments or generations. Finally, we found that a training population consisting of 100-120 plants was sufficient to train a model for accurate prediction of a quantitative trait using the genes controlling the trait. Therefore, the genes controlling a quantitative trait are capable of accurately predicting its phenotype, thereby dramatically improving the ability, accuracy, and efficiency of phenotype prediction and promoting gene-based breeding in cotton and other species.

Bayesian gamma-negative binomial modeling of single-cell RNA sequencing data.

BACKGROUND: Single-cell RNA sequencing (scRNA-seq) is a powerful profiling technique at the single-cell resolution. Appropriate analysis of scRNA-seq data can characterize molecular heterogeneity and shed light into the underlying cellular process to better understand development and disease mechanisms. The unique analytic challenge is to appropriately model highly over-dispersed scRNA-seq count data with prevalent dropouts (zero counts), making zero-inflated dimensionality reduction techniques popular for scRNA-seq data analyses. Employing zero-inflated distributions, however, may place extra emphasis on zero counts, leading to potential bias when identifying the latent structure of the data. RESULTS: In this paper, we propose a fully generative hierarchical gamma-negative binomial (hGNB) model of scRNA-seq data, obviating the need for explicitly modeling zero inflation. At the same time, hGNB can naturally account for covariate effects at both the gene and cell levels to identify complex latent representations of scRNA-seq data, without the need for commonly adopted pre-processing steps such as normalization. Efficient Bayesian model inference is derived by exploiting conditional conjugacy via novel data augmentation techniques. CONCLUSION: Experimental results on both simulated data and several real-world scRNA-seq datasets suggest that hGNB is a powerful tool for cell cluster discovery as well as cell lineage inference.

Universal promoter scanning by Pol II during transcription initiation in Saccharomyces cerevisiae.

BACKGROUND: The majority of eukaryotic promoters utilize multiple transcription start sites (TSSs). How multiple TSSs are specified at individual promoters across eukaryotes is not understood for most species. In Saccharomyces cerevisiae, a pre-initiation complex (PIC) comprised of Pol II and conserved general transcription factors (GTFs) assembles and opens DNA upstream of TSSs. Evidence from model promoters indicates that the PIC scans from upstream to downstream to identify TSSs. Prior results suggest that TSS distributions at promoters where scanning occurs shift in a polar fashion upon alteration in Pol II catalytic activity or GTF function. RESULTS: To determine the extent of promoter scanning across promoter classes in S. cerevisiae, we perturb Pol II catalytic activity and GTF function and analyze their effects on TSS usage genome-wide. We find that alterations to Pol II, TFIIB, or TFIIF function widely alter the initiation landscape consistent with promoter scanning operating at all yeast promoters, regardless of promoter class. Promoter architecture, however, can determine the extent of promoter sensitivity to altered Pol II activity in ways that are predicted by a scanning model. CONCLUSIONS: Our observations coupled with previous data validate key predictions of the scanning model for Pol II initiation in yeast, which we term the shooting gallery. In this model, Pol II catalytic activity and the rate and processivity of Pol II scanning together with promoter sequence determine the distribution of TSSs and their usage.

Accurate prediction of maize grain yield using its contributing genes for gene-based breeding.

Accurately predicting the phenotypes of complex traits is crucial to enhanced breeding in plants and livestock, and to enhanced medicine in humans. Here we reports the first study accurately predicting complex traits using their contributing genes, especially their number of favorable alleles (NFAs), genotypes and transcript expressions, with the grain yield of maize, Zea mays L. When the NFAs or genotypes of only 27 SNP/InDel-containing grain yield genes were used, a prediction accuracy of r = 0.52 or 0.49 was obtained. When the expressions of grain yield gene transcripts were used, a plateaued prediction accuracy of r = 0.84 was achieved. When the phenotypes predicted with two or three of the genic datasets were used for progeny selection, the selected lines were completely consistent with those selected by phenotypic selection. Therefore, the genes controlling complex traits enable accurately predicting their phenotypes, thus desirable for gene-based breeding in crop plants.

De novo transcriptome sequencing and analysis revealed the molecular basis of rapid fat accumulation by black soldier fly (Hermetia illucens, L.) for development of insectival biodiesel.

BACKGROUND: Black soldier fly (BSF, Hermetia illucens L.) can efficiently degrade organic wastes and transform into a high fat containing insect biomass that could be used as feedstock for biodiesel production. Meanwhile, the molecular regulatory basis of fat accumulation by BSF is still unclear; it is necessary to identify vital genes and regulators that are involved in fat accumulation. RESULTS: This study analyzed the dynamic state of fat content and fatty-acid composition of BSF larvae in eight different stages. The late prepupa stage exhibited the highest crude fat, with lauric acid being the main component. Therefore, to provide insight into this unexplained phenomenon, the molecular regulation of rapid fat accumulation by BSF larvae was investigated. The twelve developmental stages of BSF were selected for transcriptome analysis, including the eight stages used for investigation of fat content and fatty-acid composition. By Illumina sequencing, 218,295,450,000 nt were generated. Through assembly by Trinity, 70,475 unigenes were obtained with an average length of 1064 nt and an N50 of 1749 nt. The differentially expressed unigenes were identified by DESeq, with 9159 of them being up-regulated and 10,101 of them were down-regulated. The several putative genes that are involved in the formation of pyruvate, acetyl-CoA biosynthesis, acetyl-CoA transcription, fatty-acid biosynthesis, and triacylglycerol biosynthesis were identified. The four vital metabolic genes that are associated with fat accumulation were validated by quantitative real-time PCR (qRT-PCR). The molecular mechanism of fat accumulation in BSF was clarified in this investigation through the construction of a detailed fat accumulation model from our results. CONCLUSION: The study provides an unprecedented level of insight from transcriptome sequencing to reveal the crude fat accumulation mechanism in developing BSF. The finding holds considerable promise for insectival biodiesel production, and the fat content and fatty-acid composition can be altered by genetic engineering approaches in the future for the insect production industry.

Gene expression correlates of facultative predation in the blow fly Chrysomya rufifacies (Diptera: Calliphoridae).

Effects of intraguild predation (IGP) on omnivores and detritivores are relatively understudied when compared to work on predator guilds. Functional genetic work in IGP is even more limited, but its application can help answer a range of questions related to ultimate and proximate causes of this behavior. Here, we integrate behavioral assays and transcriptomic analysis of facultative predation in a blow fly (Diptera: Calliphoridae) to evaluate the prevalence, effect, and correlated gene expression of facultative predation by the invasive species Chrysomya rufifacies. Field work observing donated human cadavers indicated facultative predation by C. rufifacies on the native blow fly Cochliomyia macellaria was rare under undisturbed conditions, owing in part to spatial segregation between species. Laboratory assays under conditions of starvation showed predation had a direct fitness benefit (i.e., survival) to the predator. As a genome is not available for C. rufifacies, a de novo transcriptome was developed and annotated using sequence similarity to Drosophila melanogaster. Under a variety of assembly parameters, several genes were identified as being differentially expressed between predators and nonpredators of this species, including genes involved in cell-to-cell signaling, osmotic regulation, starvation responses, and dopamine regulation. Results of this work were integrated to develop a model of the processes and genetic regulation controlling facultative predation.

Identifying similar transcripts in a related organism from de Bruijn graphs of RNA-Seq data, with applications to the study of salt and waterlogging tolerance in Melilotus.

BACKGROUND: A popular strategy to study alternative splicing in non-model organisms starts from sequencing the entire transcriptome, then assembling the reads by using de novo transcriptome assembly algorithms to obtain predicted transcripts. A similarity search algorithm is then applied to a related organism to infer possible function of these predicted transcripts. While some of these predictions may be inaccurate and transcripts with low coverage are often missed, we observe that it is possible to obtain a more complete set of transcripts to facilitate possible functional assignments by starting the search from the intermediate de Bruijn graph that contains all branching possibilities. RESULTS: We develop an algorithm to extract similar transcripts in a related organism by starting the search from the de Bruijn graph that represents the transcriptome instead of from predicted transcripts. We show that our algorithm is able to recover more similar transcripts than existing algorithms, with large improvements in obtaining longer transcripts and a finer resolution of isoforms. We apply our algorithm to study salt and waterlogging tolerance in two Melilotus species by constructing new RNA-Seq libraries. CONCLUSIONS: We have developed an algorithm to identify paths in the de Bruijn graph that correspond to similar transcripts in a related organism directly. Our strategy bypasses the transcript prediction step in RNA-Seq data and makes use of support from evolutionary information.

Machine learning performance in a microbial molecular autopsy context: A cross-sectional postmortem human population study.

BACKGROUND: The postmortem microbiome can provide valuable information to a death investigation and to the human health of the once living. Microbiome sequencing produces, in general, large multi-dimensional datasets that can be difficult to analyze and interpret. Machine learning methods can be useful in overcoming this analytical challenge. However, different methods employ distinct strategies to handle complex datasets. It is unclear whether one method is more appropriate than others for modeling postmortem microbiomes and their ability to predict attributes of interest in death investigations, which require understanding of how the microbial communities change after death and may represent those of the once living host. METHODS AND FINDINGS: Postmortem microbiomes were collected by swabbing five anatomical areas during routine death investigation, sequenced and analyzed from 188 death cases. Three machine learning methods (boosted algorithms, random forests, and neural networks) were compared with respect to their abilities to predict case attributes: postmortem interval (PMI), location of death, and manner of death. Accuracy depended on the method used, the numbers of anatomical areas analyzed, and the predicted attribute of death. CONCLUSIONS: All algorithms performed well but with distinct features to their performance. Xgboost often produced the most accurate predictions but may also be more prone to overfitting. Random forest was the most stable across predictions that included more anatomic areas. Analysis of postmortem microbiota from more than three anatomic areas appears to yield limited returns on accuracy, with the eyes and rectum providing the most useful information correlating with circumstances of death in most cases for this dataset.