Genomic benchmarks: a collection of datasets for genomic sequence classification.

BACKGROUND: Recently, deep neural networks have been successfully applied in many biological fields. In 2020, a deep learning model AlphaFold won the protein folding competition with predicted structures within the error tolerance of experimental methods. However, this solution to the most prominent bioinformatic challenge of the past 50 years has been possible only thanks to a carefully curated benchmark of experimentally predicted protein structures. In Genomics, we have similar challenges (annotation of genomes and identification of functional elements) but currently, we lack benchmarks similar to protein folding competition. RESULTS: Here we present a collection of curated and easily accessible sequence classification datasets in the field of genomics. The proposed collection is based on a combination of novel datasets constructed from the mining of publicly available databases and existing datasets obtained from published articles. The collection currently contains nine datasets that focus on regulatory elements (promoters, enhancers, open chromatin region) from three model organisms: human, mouse, and roundworm. A simple convolution neural network is also included in a repository and can be used as a baseline model. Benchmarks and the baseline model are distributed as the Python package 'genomic-benchmarks', and the code is available at https://github.com/ML-Bioinfo-CEITEC/genomic_benchmarks . CONCLUSIONS: Deep learning techniques revolutionized many biological fields but mainly thanks to the carefully curated benchmarks. For the field of Genomics, we propose a collection of benchmark datasets for the classification of genomic sequences with an interface for the most commonly used deep learning libraries, implementation of the simple neural network and a training framework that can be used as a starting point for future research. The main aim of this effort is to create a repository for shared datasets that will make machine learning for genomics more comparable and reproducible while reducing the overhead of researchers who want to enter the field, leading to healthy competition and new discoveries.

Genic and chromosomal components of Prdm9-driven hybrid male sterility in mice (Mus musculus).

Hybrid sterility contributes to speciation by preventing gene flow between related taxa. Prdm9, the first and only hybrid male sterility gene known in vertebrates, predetermines the sites of recombination between homologous chromosomes and their synapsis in early meiotic prophase. The asymmetric binding of PRDM9 to heterosubspecific homologs of Mus musculus musculus × Mus musculus domesticus F1 hybrids and increase of PRDM9-independent DNA double-strand break hotspots results indificult- to- repair double-strand breaks, incomplete synapsis of homologous chromosomes, and meiotic arrest at the first meiotic prophase. Here, we show that Prdm9 behaves as a major hybrid male sterility gene in mice outside the Mus musculus musculus × Mus musculus domesticus F1 hybrids, in the genomes composed of Mus musculus castaneus and Mus musculus musculus chromosomes segregating on the Mus musculus domesticus background. The Prdm9cst/dom2 (castaneus/domesticus) allelic combination secures meiotic synapsis, testes weight, and sperm count within physiological limits, while the Prdm9msc1/dom2 (musculus/domesticus) males show a range of fertility impairment. Out of 5 quantitative trait loci contributing to the Prdm9msc1/dom2-related infertility, 4 control either meiotic synapsis or fertility phenotypes and 1 controls both, synapsis, and fertility. Whole-genome genotyping of individual chromosomes showed preferential involvement of nonrecombinant musculus chromosomes in asynapsis in accordance with the chromosomal character of hybrid male sterility. Moreover, we show that the overall asynapsis rate can be estimated solely from the genotype of individual males by scoring the effect of nonrecombinant musculus chromosomes. Prdm9-controlled hybrid male sterility represents an example of genetic architecture of hybrid male sterility consisting of genic and chromosomal components.

Chromosome-wide characterization of meiotic noncrossovers (gene conversions) in mouse hybrids.

During meiosis, the recombination-initiating DNA double-strand breaks (DSBs) are repaired by crossovers or noncrossovers (gene conversions). While crossovers are easily detectable, noncrossover identification is hampered by the small size of their converted tracts and the necessity of sequence polymorphism. We report identification and characterization of a mouse chromosome-wide set of noncrossovers by next-generation sequencing of 10 mouse intersubspecific chromosome substitution strains. Based on 94 identified noncrossovers, we determined the mean length of a conversion tract to be 32 bp. The spatial chromosome-wide distribution of noncrossovers and crossovers significantly differed, although both sets overlapped the known hotspots of PRDM9-directed histone methylation and DNA DSBs, thus supporting their origin in the standard DSB repair pathway. A significant deficit of noncrossovers descending from asymmetric DSBs proved their proposed adverse effect on meiotic recombination and pointed to sister chromatids as an alternative template for their repair. The finding has implications for the molecular mechanism of hybrid sterility in mice from crosses between closely related Mus musculus musculus and Mus musculus domesticus subspecies.

PENGUINN: Precise Exploration of Nuclear G-Quadruplexes Using Interpretable Neural Networks.

G-quadruplexes (G4s) are a class of stable structural nucleic acid secondary structures that are known to play a role in a wide spectrum of genomic functions, such as DNA replication and transcription. The classical understanding of G4 structure points to four variable length guanine strands joined by variable length nucleotide stretches. Experiments using G4 immunoprecipitation and sequencing experiments have produced a high number of highly probable G4 forming genomic sequences. The expense and technical difficulty of experimental techniques highlights the need for computational approaches of G4 identification. Here, we present PENGUINN, a machine learning method based on Convolutional neural networks, that learns the characteristics of G4 sequences and accurately predicts G4s outperforming state-of-the-art methods. We provide both a standalone implementation of the trained model, and a web application that can be used to evaluate sequences for their G4 potential.

High-throughput sleep phenotyping produces robust and heritable traits in Diversity Outbred mice and their founder strains.

STUDY OBJECTIVES: This study describes high-throughput phenotyping strategies for sleep and circadian behavior in mice, including examinations of robustness, reliability, and heritability among Diversity Outbred (DO) mice and their eight founder strains. METHODS: We performed high-throughput sleep and circadian phenotyping in male mice from the DO population (n = 338) and their eight founder strains: A/J (n = 6), C57BL/6J (n = 14), 129S1/SvlmJ (n = 6), NOD/LtJ (n = 6), NZO/H1LtJ (n = 6), CAST/EiJ (n = 8), PWK/PhJ (n = 8), and WSB/EiJ (n = 6). Using infrared beam break systems, we defined sleep as at least 40 s of continuous inactivity and quantified sleep-wake amounts and bout characteristics. We developed assays to measure sleep latency in a new environment and during a modified Murine Multiple Sleep Latency Test, and estimated circadian period from wheel-running experiments. For each trait, broad-sense heritability (proportion of variability explained by all genetic factors) was derived in founder strains, while narrow-sense heritability (proportion of variability explained by additive genetic effects) was calculated in DO mice. RESULTS: Phenotypes were robust to different inactivity durations to define sleep. Differences across founder strains and moderate/high broad-sense heritability were observed for most traits. There was large phenotypic variability among DO mice, and phenotypes were reliable, although estimates of heritability were lower than in founder mice. This likely reflects important nonadditive genetic effects. CONCLUSIONS: A high-throughput phenotyping strategy in mice, based primarily on monitoring of activity patterns, provides reliable and heritable estimates of sleep and circadian traits. This approach is suitable for discovery analyses in DO mice, where genetic factors explain some proportion of phenotypic variation.

R/qtl2: Software for Mapping Quantitative Trait Loci with High-Dimensional Data and Multiparent Populations.

R/qtl2 is an interactive software environment for mapping quantitative trait loci (QTL) in experimental populations. The R/qtl2 software expands the scope of the widely used R/qtl software package to include multiparent populations derived from more than two founder strains, such as the Collaborative Cross and Diversity Outbred mice, heterogeneous stocks, and MAGIC plant populations. R/qtl2 is designed to handle modern high-density genotyping data and high-dimensional molecular phenotypes, including gene expression and proteomics. R/qtl2 includes the ability to perform genome scans using a linear mixed model to account for population structure, and also includes features to impute SNPs based on founder strain genomes and to carry out association mapping. The R/qtl2 software provides all of the basic features needed for QTL mapping, including graphical displays and summary reports, and it can be extended through the creation of add-on packages. R/qtl2, which is free and open source software written in the R and C++ programming languages, comes with a test framework.

Genetic Drivers of Pancreatic Islet Function.

The majority of gene loci that have been associated with type 2 diabetes play a role in pancreatic islet function. To evaluate the role of islet gene expression in the etiology of diabetes, we sensitized a genetically diverse mouse population with a Western diet high in fat (45% kcal) and sucrose (34%) and carried out genome-wide association mapping of diabetes-related phenotypes. We quantified mRNA abundance in the islets and identified 18,820 expression QTL. We applied mediation analysis to identify candidate causal driver genes at loci that affect the abundance of numerous transcripts. These include two genes previously associated with monogenic diabetes (PDX1 and HNF4A), as well as three genes with nominal association with diabetes-related traits in humans (FAM83E, IL6ST, and SAT2). We grouped transcripts into gene modules and mapped regulatory loci for modules enriched with transcripts specific for α-cells, and another specific for δ-cells. However, no single module enriched for ß-cell-specific transcripts, suggesting heterogeneity of gene expression patterns within the ß-cell population. A module enriched in transcripts associated with branched-chain amino acid metabolism was the most strongly correlated with physiological traits that reflect insulin resistance. Although the mice in this study were not overtly diabetic, the analysis of pancreatic islet gene expression under dietary-induced stress enabled us to identify correlated variation in groups of genes that are functionally linked to diabetes-associated physiological traits. Our analysis suggests an expected degree of concordance between diabetes-associated loci in the mouse and those found in human populations, and demonstrates how the mouse can provide evidence to support nominal associations found in human genome-wide association mapping.

High-Resolution Maps of Mouse Reference Populations.

Genetic reference panels are widely used to map complex, quantitative traits in model organisms. We have generated new high-resolution genetic maps of 259 mouse inbred strains from recombinant inbred strain panels (C57BL/6J × DBA/2J, ILS/IbgTejJ × ISS/IbgTejJ, and C57BL/6J × A/J) and chromosome substitution strain panels (C57BL/6J-Chr#, C57BL/6J-Chr#, and C57BL/6J-Chr#). We genotyped all samples using the Affymetrix Mouse Diversity Array with an average intermarker spacing of 4.3 kb. The new genetic maps provide increased precision in the localization of recombination breakpoints compared to the previous maps. Although the strains were presumed to be fully inbred, we found residual heterozygosity in 40% of individual mice from five of the six panels. We also identified de novo deletions and duplications, in homozygous or heterozygous state, ranging in size from 21 kb to 8.4 Mb. Almost two-thirds (46 out of 76) of these deletions overlap exons of protein coding genes and may have phenotypic consequences. Twenty-nine putative gene conversions were identified in the chromosome substitution strains. We find that gene conversions are more likely to occur in regions where the homologous chromosomes are more similar. The raw genotyping data and genetic maps of these strain panels are available at http://churchill-lab.jax.org/website/MDA.

Defining the consequences of genetic variation on a proteome-wide scale.

Genetic variation modulates protein expression through both transcriptional and post-transcriptional mechanisms. To characterize the consequences of natural genetic diversity on the proteome, here we combine a multiplexed, mass spectrometry-based method for protein quantification with an emerging outbred mouse model containing extensive genetic variation from eight inbred founder strains. By measuring genome-wide transcript and protein expression in livers from 192 Diversity outbred mice, we identify 2,866 protein quantitative trait loci (pQTL) with twice as many local as distant genetic variants. These data support distinct transcriptional and post-transcriptional models underlying the observed pQTL effects. Using a sensitive approach to mediation analysis, we often identified a second protein or transcript as the causal mediator of distant pQTL. Our analysis reveals an extensive network of direct protein-protein interactions. Finally, we show that local genotype can provide accurate predictions of protein abundance in an independent cohort of collaborative cross mice.

Genetic Architectures of Quantitative Variation in RNA Editing Pathways.

RNA editing refers to post-transcriptional processes that alter the base sequence of RNA. Recently, hundreds of new RNA editing targets have been reported. However, the mechanisms that determine the specificity and degree of editing are not well understood. We examined quantitative variation of site-specific editing in a genetically diverse multiparent population, Diversity Outbred mice, and mapped polymorphic loci that alter editing ratios globally for C-to-U editing and at specific sites for A-to-I editing. An allelic series in the C-to-U editing enzyme Apobec1 influences the editing efficiency of Apob and 58 additional C-to-U editing targets. We identified 49 A-to-I editing sites with polymorphisms in the edited transcript that alter editing efficiency. In contrast to the shared genetic control of C-to-U editing, most of the variable A-to-I editing sites were determined by local nucleotide polymorphisms in proximity to the editing site in the RNA secondary structure. Our results indicate that RNA editing is a quantitative trait subject to genetic variation and that evolutionary constraints have given rise to distinct genetic architectures in the two canonical types of RNA editing.

High-Density Genotypes of Inbred Mouse Strains: Improved Power and Precision of Association Mapping.

Human genome-wide association studies have identified thousands of loci associated with disease phenotypes. Genome-wide association studies also have become feasible using rodent models and these have some important advantages over human studies, including controlled environment, access to tissues for molecular profiling, reproducible genotypes, and a wide array of techniques for experimental validation. Association mapping with common mouse inbred strains generally requires 100 or more strains to achieve sufficient power and mapping resolution; in contrast, sample sizes for human studies typically are one or more orders of magnitude greater than this. To enable well-powered studies in mice, we have generated high-density genotypes for ∼175 inbred strains of mice using the Mouse Diversity Array. These new data increase marker density by 1.9-fold, have reduced missing data rates, and provide more accurate identification of heterozygous regions compared with previous genotype data. We report the discovery of new loci from previously reported association mapping studies using the new genotype data. The data are freely available for download, and Web-based tools provide easy access for association mapping and viewing of the underlying intensity data for individual loci.

Genetic Analysis of Substrain Divergence in Non-Obese Diabetic (NOD) Mice.

The non-obese diabetic (NOD) mouse is a polygenic model for type 1 diabetes that is characterized by insulitis, a leukocytic infiltration of the pancreatic islets. During ~35 years since the original inbred strain was developed in Japan, NOD substrains have been established at different laboratories around the world. Although environmental differences among NOD colonies capable of impacting diabetes incidence have been recognized, differences arising from genetic divergence have not been analyzed previously. We use both mouse diversity array and whole-exome capture sequencing platforms to identify genetic differences distinguishing five NOD substrains. We describe 64 single-nucleotide polymorphisms, and two short indels that differ in coding regions of the five NOD substrains. A 100-kb deletion on Chromosome 3 distinguishes NOD/ShiLtJ and NOD/ShiLtDvs from three other substrains, whereas a 111-kb deletion in the Icam2 gene on Chromosome 11 is unique to the NOD/ShiLtDvs genome. The extent of genetic divergence for NOD substrains is compared with similar studies for C57BL6 and BALB/c substrains. As mutations are fixed to homozygosity by continued inbreeding, significant differences in substrain phenotypes are to be expected. These results emphasize the importance of using embryo freezing methods to minimize genetic drift within substrains and of applying appropriate genetic nomenclature to permit substrain recognition when one is used.

A multi-megabase copy number gain causes maternal transmission ratio distortion on mouse chromosome 2.

Significant departures from expected Mendelian inheritance ratios (transmission ratio distortion, TRD) are frequently observed in both experimental crosses and natural populations. TRD on mouse Chromosome (Chr) 2 has been reported in multiple experimental crosses, including the Collaborative Cross (CC). Among the eight CC founder inbred strains, we found that Chr 2 TRD was exclusive to females that were heterozygous for the WSB/EiJ allele within a 9.3 Mb region (Chr 2 76.9 - 86.2 Mb). A copy number gain of a 127 kb-long DNA segment (designated as responder to drive, R2d) emerged as the strongest candidate for the causative allele. We mapped R2d sequences to two loci within the candidate interval. R2d1 is located near the proximal boundary, and contains a single copy of R2d in all strains tested. R2d2 maps to a 900 kb interval, and the number of R2d copies varies from zero in classical strains (including the mouse reference genome) to more than 30 in wild-derived strains. Using real-time PCR assays for the copy number, we identified a mutation (R2d2WSBdel1) that eliminates the majority of the R2d2WSB copies without apparent alterations of the surrounding WSB/EiJ haplotype. In a three-generation pedigree segregating for R2d2WSBdel1, the mutation is transmitted to the progeny and Mendelian segregation is restored in females heterozygous for R2d2WSBdel1, thus providing direct evidence that the copy number gain is causal for maternal TRD. We found that transmission ratios in R2d2WSB heterozygous females vary between Mendelian segregation and complete distortion depending on the genetic background, and that TRD is under genetic control of unlinked distorter loci. Although the R2d2WSB transmission ratio was inversely correlated with average litter size, several independent lines of evidence support the contention that female meiotic drive is the cause of the distortion. We discuss the implications and potential applications of this novel meiotic drive system.

Quantitative trait locus mapping methods for diversity outbred mice.

Genetic mapping studies in the mouse and other model organisms are used to search for genes underlying complex phenotypes. Traditional genetic mapping studies that employ single-generation crosses have poor mapping resolution and limit discovery to loci that are polymorphic between the two parental strains. Multiparent outbreeding populations address these shortcomings by increasing the density of recombination events and introducing allelic variants from multiple founder strains. However, multiparent crosses present new analytical challenges and require specialized software to take full advantage of these benefits. Each animal in an outbreeding population is genetically unique and must be genotyped using a high-density marker set; regression models for mapping must accommodate multiple founder alleles, and complex breeding designs give rise to polygenic covariance among related animals that must be accounted for in mapping analysis. The Diversity Outbred (DO) mice combine the genetic diversity of eight founder strains in a multigenerational breeding design that has been maintained for >16 generations. The large population size and randomized mating ensure the long-term genetic stability of this population. We present a complete analytical pipeline for genetic mapping in DO mice, including algorithms for probabilistic reconstruction of founder haplotypes from genotyping array intensity data, and mapping methods that accommodate multiple founder haplotypes and account for relatedness among animals. Power analysis suggests that studies with as few as 200 DO mice can detect loci with large effects, but loci that account for <5% of trait variance may require a sample size of up to 1000 animals. The methods described here are implemented in the freely available R package DOQTL.

X chromosome control of meiotic chromosome synapsis in mouse inter-subspecific hybrids.

Hybrid sterility (HS) belongs to reproductive isolation barriers that safeguard the integrity of species in statu nascendi. Although hybrid sterility occurs almost universally among animal and plant species, most of our current knowledge comes from the classical genetic studies on Drosophila interspecific crosses or introgressions. With the house mouse subspecies Mus m. musculus and Mus m. domesticus as a model, new research tools have become available for studies of the molecular mechanisms and genetic networks underlying HS. Here we used QTL analysis and intersubspecific chromosome substitution strains to identify a 4.7 Mb critical region on Chromosome X (Chr X) harboring the Hstx2 HS locus, which causes asymmetrical spermatogenic arrest in reciprocal intersubspecific F1 hybrids. Subsequently, we mapped autosomal loci on Chrs 3, 9 and 13 that can abolish this asymmetry. Combination of immunofluorescent visualization of the proteins of synaptonemal complexes with whole-chromosome DNA FISH on pachytene spreads revealed that heterosubspecific, unlike consubspecific, homologous chromosomes are predisposed to asynapsis in F1 hybrid male and female meiosis. The asynapsis is under the trans- control of Hstx2 and Hst1/Prdm9 hybrid sterility genes in pachynemas of male but not female hybrids. The finding concurred with the fertility of intersubpecific F1 hybrid females homozygous for the Hstx2(Mmm) allele and resolved the apparent conflict with the dominance theory of Haldane's rule. We propose that meiotic asynapsis in intersubspecific hybrids is a consequence of cis-acting mismatch between homologous chromosomes modulated by the trans-acting Hstx2 and Prdm9 hybrid male sterility genes.

Encoding of situations in the vocal repertoire of piglets (Sus scrofa): a comparison of discrete and graded classifications.

Two important questions in bioacoustics are whether vocal repertoires of animals are graded or discrete and how the vocal expressions are linked to the context of emission. Here we address these questions in an ungulate species. The vocal repertoire of young domestic pigs, Sus scrofa, was quantitatively described based on 1513 calls recorded in 11 situations. We described the acoustic quality of calls with 8 acoustic parameters. Based on these parameters, the k-means clustering method showed a possibility to distinguish either two or five clusters although the call types are rather blurred than strictly discrete. The division of the vocal repertoire of piglets into two call types has previously been used in many experimental studies into pig acoustic communication and the five call types correspond well to previously published partial repertoires in specific situations. Clear links exist between the type of situation, its putative valence, and the vocal expression in that situation. These links can be described adequately both with a set of quantitative acoustic variables and through categorisation into call types. The information about the situation of emission of the calls is encoded through five call types almost as accurately as through the full quantitative description.

Mechanistic basis of infertility of mouse intersubspecific hybrids.

According to the Dobzhansky-Muller model, hybrid sterility is a consequence of the independent evolution of related taxa resulting in incompatible genomic interactions of their hybrids. The model implies that the incompatibilities evolve randomly, unless a particular gene or nongenic sequence diverges much faster than the rest of the genome. Here we propose that asynapsis of heterospecific chromosomes in meiotic prophase provides a recurrently evolving trigger for the meiotic arrest of interspecific F1 hybrids. We observed extensive asynapsis of chromosomes and disturbance of the sex body in >95% of pachynemas of Mus m. musculus × Mus m. domesticus sterile F1 males. Asynapsis was not preceded by a failure of double-strand break induction, and the rate of meiotic crossing over was not affected in synapsed chromosomes. DNA double-strand break repair was delayed or failed in unsynapsed autosomes, and misexpression of chromosome X and chromosome Y genes was detected in single pachynemas and by genome-wide expression profiling. Oocytes of F1 hybrid females showed the same kind of synaptic problems but with the incidence reduced to half. Most of the oocytes with pachytene asynapsis were eliminated before birth. We propose the heterospecific pairing of homologous chromosomes as a preexisting condition of asynapsis in interspecific hybrids. The asynapsis may represent a universal mechanistic basis of F1 hybrid sterility manifested by pachytene arrest. It is tempting to speculate that a fast-evolving subset of the noncoding genomic sequence important for chromosome pairing and synapsis may be the culprit.

Interallelic and intergenic incompatibilities of the Prdm9 (Hst1) gene in mouse hybrid sterility.

The Dobzhansky-Muller model of incompatibilities explains reproductive isolation between species by incorrect epistatic interactions. Although the mechanisms of speciation are of great interest, no incompatibility has been characterized at the gene level in mammals. The Hybrid sterility 1 gene (Hst1) participates in the arrest of meiosis in F(1) males of certain strains from two Mus musculus subspecies, e.g., PWD from M. m. musculus and C57BL/6J (henceforth B6) from M. m. domesticus. Hst1 has been identified as a meiotic PR-domain gene (Prdm9) encoding histone 3 methyltransferase in the male offspring of PWD females and B6 males, (PWD×B6)F(1). To characterize the incompatibilities underlying hybrid sterility, we phenotyped reproductive and meiotic markers in males with altered copy numbers of Prdm9. A partial rescue of fertility was observed upon removal of the B6 allele of Prdm9 from the azoospermic (PWD×B6)F(1) hybrids, whereas removing one of the two Prdm9 copies in PWD or B6 background had no effect on male reproduction. Incompatibility(ies) not involving Prdm9(B6) also acts in the (PWD×B6)F(1) hybrids, since the correction of hybrid sterility by Prdm9(B6) deletion was not complete. Additions and subtractions of Prdm9 copies, as well as allelic replacements, improved meiotic progression and fecundity also in the progeny-producing reciprocal (B6×PWD)F(1) males. Moreover, an increased dosage of Prdm9 and reciprocal cross enhanced fertility of other sperm-carrying male hybrids, (PWD×B6-C3H.Prdm9)F(1), harboring another Prdm9 allele of M. m. domesticus origin. The levels of Prdm9 mRNA isoforms were similar in the prepubertal testes of all types of F(1) hybrids of PWD with B6 and B6-C3H.Prdm9 despite their different prospective fertility, but decreased to 53% after removal of Prdm9(B6). Therefore, the Prdm9(B6) allele probably takes part in posttranscriptional dominant-negative hybrid interaction(s) absent in the parental strains.

Dissecting the genetic architecture of F1 hybrid sterility in house mice.

Hybrid sterility as a postzygotic reproductive isolation mechanism has been studied for over 80 years, yet the first identifications of hybrid sterility genes in Drosophila and mouse are quite recent. To study the genetic architecture of F(1) hybrid sterility between young subspecies of house mouse Mus m. domesticus and M. m. musculus, we conducted QTL analysis of a backcross between inbred strains representing these two subspecies and probed the role of individual chromosomes in hybrid sterility using the intersubspecific chromosome substitution strains. We provide direct evidence that the asymmetry in male infertility between reciprocal crosses is conferred by the middle region of M. m. musculus Chr X, thus excluding other potential candidates such as Y, imprinted genes, and mitochondrial DNA. QTL analysis identified strong hybrid sterility loci on Chr 17 and Chr X and predicted a set of interchangeable autosomal loci, a subset of which is sufficient to activate the Dobzhansky-Muller incompatibility of the strong loci. Overall, our results indicate the oligogenic nature of F(1) hybrid sterility, which should be amenable to reconstruction by proper combination of chromosome substitution strains. Such a prefabricated model system should help to uncover the gene networks and molecular mechanisms underlying hybrid sterility.