BACKGROUND: Methods for inferring the three-dimensional (3D) configuration of chromatin from conformation capture assays that provide strictly pairwise interactions, notably Hi-C, utilize the attendant contact matrix as input. More recent assays, in particular split-pool recognition of interactions by tag extension (SPRITE), capture multi-way interactions instead of solely pairwise contacts. These assays yield contacts that straddle appreciably greater genomic distances than Hi-C, in addition to instances of exceptionally high-order chromatin interaction. Such attributes are anticipated to be consequential with respect to 3D genome reconstruction, a task yet to be undertaken with multi-way contact data. However, performing such 3D reconstruction using distance-based reconstruction techniques requires framing multi-way contacts as (pairwise) distances. Comparing approaches for so doing, and assessing the resultant impact of long-range and multi-way contacts, are the objectives of this study. RESULTS: We obtained 3D reconstructions via multi-dimensional scaling under a variety of weighting schemes for mapping SPRITE multi-way contacts to pairwise distances. Resultant configurations were compared following Procrustes alignment and relationships were assessed between associated Procrustes root mean square errors and key features such as the extent of multi-way and/or long-range contacts. We found that these features had surprisingly limited influence on 3D reconstruction, a finding we attribute to their influence being diminished by the preponderance of pairwise contacts. CONCLUSION: Distance-based 3D genome reconstruction using SPRITE multi-way contact data is not appreciably affected by the weighting scheme used to convert multi-way interactions to pairwise distances.

BACKGROUND: The spatial configuration of chromosomes is essential to various cellular processes, notably gene regulation, while architecture related alterations, such as translocations and gene fusions, are often cancer drivers. Thus, eliciting chromatin conformation is important, yet challenging due to compaction, dynamics and scale. However, a variety of recent assays, in particular Hi-C, have generated new details of chromatin structure, spawning a number of novel biological findings. Many findings have resulted from analyses on the level of native contact data as generated by the assays. Alternatively, reconstruction based approaches often proceed by first converting contact frequencies into distances, then generating a three dimensional (3D) chromatin configuration that best recapitulates these distances. Subsequent analyses can enrich contact level analyses via superposition of genomic attributes on the reconstruction. But, such advantages depend on the accuracy of the reconstruction which, absent gold standards, is inherently difficult to assess. Attempts at accuracy evaluation have relied on simulation and/or FISH imaging that typically features a handful of low resolution probes. While newly advanced multiplexed FISH imaging offers possibilities for refined 3D reconstruction accuracy evaluation, availability of such data is limited due to assay complexity and the resolution thereof is appreciably lower than the reconstructions being assessed. Accordingly, there is demand for new methods of reconstruction accuracy appraisal. RESULTS: Here we explore the potential of recently proposed stationary distributions, hereafter StatDns, derived from Hi-C contact matrices, to serve as a basis for reconstruction accuracy assessment. Current usage of such StatDns has focussed on the identification of highly interactive regions (HIRs): computationally defined regions of the genome purportedly involved in numerous long-range intra-chromosomal contacts. Consistent identification of HIRs would be informative with respect to inferred 3D architecture since the corresponding regions of the reconstruction would have an elevated number of k nearest neighbors (kNNs). More generally, we anticipate a monotone decreasing relationship between StatDn values and kNN distances. After initially evaluating the reproducibility of StatDns across replicate Hi-C data sets, we use this implied StatDn - kNN relationship to gauge the utility of StatDns for reconstruction validation, making recourse to both real and simulated examples. CONCLUSIONS: Our analyses demonstrate that, as constructed, StatDns do not provide a suitable measure for assessing the accuracy of 3D genome reconstructions. Whether this is attributable to specific choices surrounding normalization in defining StatDns or to the logic underlying their very formulation remains to be determined.

BACKGROUND: Three dimensional (3D) genome spatial organization is critical for numerous cellular functions, including transcription, while certain conformation-driven structural alterations are frequently oncogenic. Genome conformation had been difficult to elucidate but the advent chromatin conformation capture assays, notably Hi-C, has transformed understanding of chromatin architecture and yielded numerous biological insights. Although most of these findings have flowed from analysis of proximity data produced by these assays, added value in generating 3D reconstructions has been demonstrated, deriving, in part, from superposing genomic features on the reconstruction. However, advantages of 3D structure-based analyses are clearly conditional on the accuracy of the attendant reconstructions, which is difficult to assess. Proponents of competing reconstruction algorithms have evaluated their accuracy by recourse to simulation of toy structures and/or limited fluorescence in situ hybridization (FISH) imaging that features a handful of low resolution probes. Accordingly, new methods of reconstruction accuracy assessment are needed. RESULTS: Here we utilize two recently devised assays to develop methodology for assessing 3D reconstruction accuracy. Multiplex FISH increases the number of probes by an order of magnitude and hence the number of inter-probe distances by two orders, providing sufficient information for structure-level evaluation via mean-squared deviations (MSD). Crucially, underscoring multiplex FISH applications are large numbers of coordinate-system aligned replicates that provide the basis for a referent distribution for MSD statistics. Using this system we show that reconstructions based on Hi-C data for IMR90 cells are accurate for some chromosomes but not others. The second new assay, genome architecture mapping, utilizes large numbers of thin cryosections to obtain a measure of proximity. We exploit the planarity of the cryosections - not used in inferring proximity - to obtain measures of reconstruction accuracy, with referents provided via resampling. Application to mouse embryonic stem cells shows reconstruction accuracies that vary by chromosome. CONCLUSIONS: We have developed methods for assessing the accuracy of 3D genome reconstructions that exploit features of recently advanced multiplex FISH and genome architecture mapping assays. These approaches can help overcome the absence of gold standards for making such assessments which are important in view of the considerable uncertainties surrounding 3D genome reconstruction.

BACKGROUND: The three-dimensional (3D) configuration of chromosomes within the eukaryote nucleus is an important factor for several cellular functions, including gene expression regulation, and has also been linked with cancer-causing translocation events. While visualization of such architecture remains limited to low resolutions, the ability to infer structures at increasing resolutions has been enabled by recently-devised chromosome conformation capture techniques. In particular, when coupled with next generation sequencing, such methods yield an inventory of genome-wide chromatin contacts or interactions. Various algorithms have been advanced to operate on such contact data to produce reconstructed 3D configurations. Studies have shown that these reconstructions can provide added value over raw interaction data with respect to downstream biological insights. However, only limited, low-resolution reconstructions have been realized for mammals due to computational bottlenecks. RESULTS: Here we propose a two-stage algorithm to partially overcome these computational barriers. The central idea is to initially utilize existing reconstruction techniques on an individual chromosome basis, using intra-chromosomal contacts, and then to relatively position these chromosome-level reconstructions using inter-chromosomal contacts. This two-stage strategy represents a natural approach in view of the within- versus between- chromosome distribution of contacts. It can increase resolution ≈ 20 fold for mouse and human. After describing the algorithm we present 3D architectures for mouse embryonic stem cells and human lymphoblastoid cells. We evaluate the impact of several factors on reconstruction reproducibility and explore a variety of sampling schemes. We further analyze replicate data at differing resolutions obtained from recently devised in situ Hi-C assays. In all instances we demonstrate insensitivity of the whole-genome 3D reconstruction obtained by the two-stage algorithm to the sampling strategy used. CONCLUSIONS: Our two-stage algorithm has the potential to significantly increase the resolution of 3D genome reconstructions. The improvements are such that we can progress from 1 Mb resolution to 100 kb resolution, notable since this latter value has been identified as critical to inferring topological domains in analyses performed on the contact (rather than 3D) level.

The recent availability of whole-genome scale data sets that investigate complementary and diverse aspects of transcriptional regulation has spawned an increased need for new and effective computational approaches to analyze and integrate these large scale assays. Here, we propose a novel algorithm, based on random forest methodology, to relate gene expression (as derived from expression microarrays) to sequence features residing in gene promoters (as derived from DNA motif data) and transcription factor binding to gene promoters (as derived from tiling microarrays). We extend the random forest approach to model a multivariate response as represented, for example, by time-course gene expression measures. An analysis of the multivariate random forest output reveals complex regulatory networks, which consist of cohesive, condition-dependent regulatory cliques. Each regulatory clique features homogeneous gene expression profiles and common motifs or synergistic motif groups. We apply our method to several yeast physiological processes: cell cycle, sporulation, and various stress conditions. Our technique displays excellent performance with regard to identifying known regulatory motifs, including high order interactions. In addition, we present evidence of the existence of an alternative MCB-binding pathway, which we confirm using data from two independent cell cycle studies and two other physioloigical processes. Finally, we have uncovered elaborate transcription regulation refinement mechanisms involving PAC and mRRPE motifs that govern essential rRNA processing. These include intriguing instances of differing motif dosages and differing combinatorial motif control that promote regulatory specificity in rRNA metabolism under differing physiological processes.

Abstract Contact: [email protected]

MOTIVATION: In the genomics setting, an increasingly common data configuration consists of a small set of sequences possessing a targeted property (positive instances) amongst a large set of sequences for which class membership is unknown (unlabeled instances). Traditional two-class classification methods do not effectively handle such data. RESULTS: Here, we develop a novel method, likely positive-iterative classification (LP-IC) for this problem, and contrast its performance with the few existing methods, most of which were devised and utilized in the text classification context. LP-IC employs an iterative classification scheme and introduces a class dispersion measure, adopted from unsupervised clustering approaches, to monitor the model selection process. Using two case studies--prediction of HLA binding, and alternative splicing conservation between human and mouse--we show that LP-IC provides superior performance to existing methodologies in terms of: (i) combined accuracy and precision in positive identification from the unlabeled set; and (ii) predictive performance of the resultant classifiers on independent test data.

MOTIVATION: Modern strategies for mapping disease loci require efficient genotyping of a large number of known polymorphic sites in the genome. The sensitive and high-throughput nature of hybridization-based DNA microarray technology provides an ideal platform for such an application by interrogating up to hundreds of thousands of single nucleotide polymorphisms (SNPs) in a single assay. Similar to the development of expression arrays, these genotyping arrays pose many data analytic challenges that are often platform specific. Affymetrix SNP arrays, e.g. use multiple sets of short oligonucleotide probes for each known SNP, and require effective statistical methods to combine these probe intensities in order to generate reliable and accurate genotype calls. RESULTS: We developed an integrated multi-SNP, multi-array genotype calling algorithm for Affymetrix SNP arrays, MAMS, that combines single-array multi-SNP (SAMS) and multi-array, single-SNP (MASS) calls to improve the accuracy of genotype calls, without the need for training data or computation-intensive normalization procedures as in other multi-array methods. The algorithm uses resampling techniques and model-based clustering to derive single array based genotype calls, which are subsequently refined by competitive genotype calls based on (MASS) clustering. The resampling scheme caps computation for single-array analysis and hence is readily scalable, important in view of expanding numbers of SNPs per array. The MASS update is designed to improve calls for atypical SNPs, harboring allele-imbalanced binding affinities, that are difficult to genotype without information from other arrays. Using a publicly available data set of HapMap samples from Affymetrix, and independent calls by alternative genotyping methods from the HapMap project, we show that our approach performs competitively to existing methods. AVAILABILITY: R functions are available upon request from the authors.

Abstract Motivation: A common objective of microarray experiments is the detection of differential gene expression between samples obtained under different conditions. The task of identifying differentially expressed genes consists of two aspects: ranking and selection. Numerous statistics have been proposed to rank genes in order of evidence for differential expression. However, no one statistic is universally optimal and there is seldom any basis or guidance that can direct toward a particular statistic of choice. Results: Our new approach, which addresses both ranking and selection of differentially expressed genes, integrates differing statistics via a distance synthesis scheme. Using a set of (Affymetrix) spike-in datasets, in which differentially expressed genes are known, we demonstrate that our method compares favorably with the best individual statistics, while achieving robustness properties lacked by the individual statistics. We further evaluate performance on one other microarray study. Availability: The approach is implemented in an R package called DEDS, which is available for download from the Bioconductor website (http://www.bioconductor.org/). Contact: [email protected]

Splicing is an important process for regulation of gene expression in eukaryotes, and it has important functional links to other steps of gene expression. Two examples of these linkages include Ceg1, a component of the mRNA capping enzyme, and the chromatin elongation factors Spt4-5, both of which have recently been shown to play a role in the normal splicing of several genes in the yeast Saccharomyces cerevisiae. Using a genomic approach to characterize the roles of Spt4-5 in splicing, we used splicing-sensitive DNA microarrays to identify specific sets of genes that are mis-spliced in ceg1, spt4, and spt5 mutants. In the context of a complex, nested, experimental design featuring 22 dye-swap array hybridizations, comprising both biological and technical replicates, we applied five appropriate statistical models for assessing differential expression between wild-type and the mutants. To refine selection of differential expression genes, we then used a robust model-synthesizing approach, Differential Expression via Distance Synthesis, to integrate all five models. The resultant list of differentially expressed genes was then further analyzed with regard to select attributes: we found that highly transcribed genes with long introns were most sensitive to spt mutations. QPCR confirmation of differential expression was established for the limited number of genes evaluated. In this paper, we showcase splicing array technology, as well as powerful, yet general, statistical methodology for assessing differential expression, in the context of a real, complex experimental design. Our results suggest that the Spt4-Spt5 complex may help coordinate splicing with transcription under conditions that present kinetic challenges to spliceosome assembly or function.

BACKGROUND: Mutations in rpoB, the gene encoding the beta subunit of DNA-dependent RNA polymerase, are associated with rifampin resistance in Mycobacterium tuberculosis. Several studies have been conducted where minimum inhibitory concentration (MIC, which is defined as the minimum concentration of the antibiotic in a given culture medium below which bacterial growth is not inhibited) of rifampin has been measured and partial DNA sequences have been determined for rpoB in different isolates of M. tuberculosis. However, no model has been constructed to predict rifampin resistance based on sequence information alone. Such a model might provide the basis for quantifying rifampin resistance status based exclusively on DNA sequence data and thus eliminate the requirements for time consuming culturing and antibiotic testing of clinical isolates. RESULTS: Sequence data for amino acid positions 511-533 of rpoB and associated MIC of rifampin for different isolates of M. tuberculosis were taken from studies examining rifampin resistance in clinical samples from New York City and throughout Japan. We used tree-based statistical methods and random forests to generate models of the relationships between rpoB amino acid sequence and rifampin resistance. The proportion of variance explained by a relatively simple tree-based cross-validated regression model involving two amino acid positions (526 and 531) is 0.679. The first partition in the data, based on position 531, results in groups that differ one hundredfold in mean MIC (1.596 micrograms/ml and 159.676 micrograms/ml). The subsequent partition based on position 526, the most variable in this region, results in a > 354-fold difference in MIC. When considered as a classification problem (susceptible or resistant), a cross-validated tree-based model correctly classified most (0.884) of the observations and was very similar to the regression model. Random forest analysis of the MIC data as a continuous variable, a regression problem, produced a model that explained 0.861 of the variance. The random forest analysis of the MIC data as discrete classes produced a model that correctly classified 0.942 of the observations with sensitivity of 0.958 and specificity of 0.885. CONCLUSIONS: Highly accurate regression and classification models of rifampin resistance can be made based on this short sequence region. Models may be better with improved (and consistent) measurements of MIC and more sequence data.