Random subwindows and extremely randomized trees for image classification in cell biology

Background

With the improvements in biosensors and high-throughput image acquisition technologies, life science laboratories are able to perform an increasing number of experiments that involve the generation of a large amount of images at different imaging modalities/scales. It stresses the need for computer vision methods that automate image classification tasks.

Results

We illustrate the potential of our image classification method in cell biology by evaluating it on four datasets of images related to protein distributions or subcellular localizations, and red-blood cell shapes. Accuracy results are quite good without any specific pre-processing neither domain knowledge incorporation. The method is implemented in Java and available upon request for evaluation and research purpose.

Conclusion

Our method is directly applicable to any image classification problems. We foresee the use of this automatic approach as a baseline method and first try on various biological image classification problems.

     

Efficiency robust tests for mapping quantitative trait loci using extremely discordant sib pairs

In 1972, Haseman and Elston proposed a pioneering regression method for mapping quantitative trait loci using randomly selected sib pairs. Recently, the statistical power of their method was shown to be increased when extremely discordant sib pairs are ascertained. While the precise genetic model may not be known, prior information that constrains IBD probabilities is often available. We investigate properties of tests that are robust against model uncertainty and show that the power gain from further constraining IBD probabilities is marginal. The additional linkage information contained in the trait values can be incorporated by combining the Haseman-Elston regression method and a robust allele sharing test.     

Flood mapping with remote sensing and hydrochemistry: A new method to distinguish the origin of flood water during floods

River flooding is important for the ecological functioning of river floodplains. It is implicitly assumed that in many river floodplains during floods, river water is spreading all over the floodplain. We hypothesize that during flood events a spatial distribution of water types exists, which is correlated to different water sources (river water, atmospheric water and groundwater) and to the spatial distribution of vegetation types. The objective of this paper is to assess a new methodology to determine the extent of flooding and the spatial distribution of different water sources during the flood, using GPS, multispectral remote sensing and hydrochemical analyses. This methodology is applied to the Biebrza River Lower Basin, which has little human impact. Remote sensing resulted in a map distinguishing inundated areas from dry areas, which showed 85% agreement with GPS field measurements. Principal Component Analyses and Cluster Analyses on the measured water chemistry identified different water sources during the flood (river water, groundwater, rainwater) and showed the effects of human impact on water quality. River flood water dominated the entire inundation zone in the northern Lower Basin, which is narrower and steeper than the southern Lower Basin where groundwater and rainwater were significant contributors to the major part of the inundated area. Vegetation in the river flood zone is distinctly different from the rest of the floodplain. Due to mixing of ground- and rainwater, correlation analyses between vegetation and water type were not possible outside the river flood zone. The new methodology is effective in distinguishing inundated areas from dry regions and in separating river flood water from other water sources during a flood.     

Quantitative trait locus mapping for atherosclerosis susceptibility

PURPOSE OF REVIEW: Atherosclerosis is a complex trait with both environmental and genetic aspects. Although some progress has been made in defining genes associated with atherosclerosis in humans, animal models have been useful in learning about pathways and genes involved in atherogenesis. This review describes an unbiased genetic mapping method called quantitative trait locus mapping and progress in using this method to identify genes that alter atherosclerosis susceptibility in mice. RECENT FINDINGS: Approximately 10 well defined genetic loci have been described that are associated with lesion severity in diet-induced or gene knockout mouse models of atherosclerosis. Recently, two of these genetic loci were narrowed considerably by analysis of genetic recombinants within these loci. In addition, a computational method to discover quantitative trait loci has been applied to atherosclerosis. However, none of the genes responsible for these atherosclerosis quantitative trait loci has been definitively identified. The recent completion of the mouse draft genome should facilitate the task of identifying these genes. SUMMARY: Quantitative trait locus mapping studies in mouse models of atherosclerosis have defined genetic regions that alter lesion severity. The identification of the responsible genes may lead to insights into the pathogenesis of atherosclerosis as well as to candidates for human genetic association studies.     

ALTree: association detection and localization of susceptibility sites using haplotype phylogenetic trees

Finding the genes involved in complex diseases susceptibility and among those genes, localizing the variant sites explaining this susceptibility is a major goal of genetic epidemiology. In this context, haplotypic methods that use the joint information on several markers may be of particular interest. When the number of haplotypes is large, a grouping may be required. Phylogenetic trees allow such groupings of haplotypes based on their evolutionary history and may help in the detection and localization of disease susceptibility sites. In this paper, we present a new software to perform phylogeny-based association and localization analysis.     

Prospects for association-based fine mapping of a susceptibility gene for a complex disease

The potential of association studies for fine-mapping loci with common disease susceptibility alleles for complex genetic diseases in outbred populations is unclear. For a battery of tightly linked anonymous genetic markers spanning a candidate region centered around a disease locus, simulation methods based on a coalescent process with mutation, recombination, and genetic drift were used to study the spatial distribution of markers with large noncentrality parameters in a case-control study design. Simulations with a disease allele at intermediate frequency, presumably representing an old mutation, tend to exhibit the largest noncentrality parameter values at markers near the disease locus. In contrast, simulations with a disease allele at low frequency, presumably representing a young mutation, often exhibit the largest noncentrality parameter values at markers scattered over the candidate region. In the former case, sample sizes or marker densities sufficient to detect association are likely to lead to useful localization, whereas, in the latter case, localization of the disease locus within the candidate region is much less likely, regardless of the sample size or density of the map. The simulations suggest that for a single marker analysis, the simple strategy of choosing the marker with smallest associated P value to begin a laboratory search for the disease locus performs adequately for a common disease allele.     

An interictal measurement of cerebral oxygen extraction fraction in MRI-negative refractory epilepsy using quantitative susceptibility mapping

PurposeOxygen extraction fraction (OEF) can be a factor to identify brain tissue’s disability in epileptic patients. This study aimed to assess the OEF’s level measurement in refractory epileptic patients (REPs) using a quantitative susceptibility mapping (QSM) method and to determine whether the OEF parameters change.MethodsQSM-OEF maps of 26 REPs and 16 healthy subjects were acquired using 3T MRI with a 64-channel coil. Eighteen regions-of-interest (ROIs) were chosen around the cortex in one appropriate slice of the brain and the mean QSM-OEF for each ROI was obtained. The correlations of QSM-OEF among different clinical characteristics of the disease, as well as between the patients and normal subjects, were also investigated.ResultsQSM-OEF was shown to be significantly higher in REPs (44.9 ± 5.8) than that in HS (41.9 ± 6.2) (p < 0.05). Mean QSM-OEF was statistically lower in the ipsilateral side (44.5 ± 6.6) compared to the contralateral side (46.4 ± 6.8) (P < 0.01). QSM-OEF was illustrated to have a strong positive correlation with the attack duration (r = 0.6), and a moderate negative correlation with the attack frequency (r = −0.3). Using an optimized support vector machine algorithm, we could predict the disease in subjects having abnormal OEF values in the brain-selected-ROIs with sensitivity, specificity, AUC, and the precision of 0.96, 1, 0.98, and 1, respectively.ConclusionsThe results of this study revealed that QSM-OEF of the REPs’ brain is higher than that of HS, which indicates that QSM-OEF is associated with disease activity.     

A nonparametric method for classification trees using grouped covariates

A group of variables are commonly seen in diagnostic medicine when multiple prognostic factors are aggregated into a composite score to represent the risk profile. A model selection method considers these covariates as all-in or all-out types. Model selection procedures for grouped covariates and their applications have thrived in recent years, in part because of the development of genetic research in which gene–gene or gene–environment interactions and regulatory network pathways are considered groups of individual variables. However, little has been discussed on how to utilize grouped covariates to grow a classification tree. In this paper, we propose a nonparametric method to address the selection of split variables for grouped covariates and their following selection of split points. Comprehensive simulations were implemented to show the superiority of our procedures compared to a commonly used recursive partition algorithm. The practical use of our method is demonstrated through a real data analysis that uses a group of prognostic factors to classify the successful mobilization of peripheral blood stem cells.     

High-resolution genome-wide linkage mapping identifies susceptibility loci for BMI in the Chinese population

The genetic loci affecting the commonly used BMI have been intensively investigated using linkage approaches in multiple populations. This study aims at performing the first genome‐wide linkage scan on BMI in the Chinese population in mainland China with hypothesis that heterogeneity in genetic linkage could exist in different ethnic populations. BMI was measured from 126 dizygotic twins in Qingdao municipality who were genotyped using high‐resolution Affymetrix Genome‐Wide Human SNP arrays containing about 1 million single‐nucleotide polymorphisms (SNPs). Nonparametric linkage analysis was performed with Merlin software package for linkage analysis using variance components approach for quantitative trait loci mapping. We identified a strong linkage peak at the end of chromosome 7 (7q36 at 186 cM) with a lod score of 4.06 which overlaps with that reported by a large multicenter study in western countries. Multiple loci showing suggestive linkage were found on chromosome 1 (lod score 2.38 at 242 cM), chromosome 8 (2.48 at 95 cM), and chromosome 14 (2.2 at 89.4 cM). The strong linkage identified in the Chinese subjects that is consistent with that found in populations of European origin could suggest the existence of evolutionarily preserved genetic mechanisms for BMI whereas the multiple suggestive loci could represent genetic effect from gene—environment interaction as a result of population‐specific environmental adaptation.     

Multi-QTL mapping for quantitative traits using distorted markers

Marker segregation distortion is a common natural phenomenon. However, relatively little is known about utilizing distorted markers for detecting quantitative trait loci (QTL). Therefore, in this study we proposed a multi-QTL mapping approach that uses distorted markers. First, the information from all markers, including distorted markers, was used to detect segregation distortion loci (SDL). Second, the information from the detected SDL was used to correct the conditional probabilities of the QTL genotypes conditional on marker information, and these corrected probabilities were then incorporated into a multi-QTL mapping methodology. Finally, the proposed approach was validated by both Monte Carlo simulation studies and real data analysis. The results from the simulation studies show that as long as one or two SDL are placed around the simulated QTL, there are no differences between the new method and the ordinary interval mapping method in terms of the power of QTL detection or the estimates of the position and dominant effects of the QTL. However, the power of QTL detection is higher under the dominant genetic model of SDL than under the additive genetic model, and the estimate for the additive effect of QTL using the new method is significantly different from the estimate obtained using ordinary interval mapping. The above results were further confirmed by the detection of QTL for dried soymilk in 222 F_2:4 families in soybean.     

Genetic mapping of genes for susceptibility to black spot disease in Japanese pears.

Black spot disease, which is caused by the Japanese pear pathotype of Alternaria alternata (Fr.) Keissler, is one of the most harmful diseases in Japanese pear cultivation. We identified the exact positions and linkage groups (LGs) of the genes for susceptibility to black spot in the Japanese pear (Pyrus pyrifolia Nakai) cultivars 'Osa Nijisseiki' (gene Ani) and 'Nansui' (gene Ana). Segregation of susceptibility and resistance fitted the expected ratio of 1:1 in progeny of 'Nansui' but showed a slight distortion in progeny of 'Osa Nijisseiki'. We mapped the genes for susceptibility to black spot in both populations using a genome scanning approach. The simple sequence repeat (SSR) markers CH04h02 and CH03d02 showed tight linkage to Ani and Ana. Although Ani and Ana are derived from different sources, both genes are located at the top region of LG 11. Information about the positions of the susceptibility genes and the molecular markers linked to them will be useful for marker-assisted selection in pear breeding programs.     

A statistical framework for eQTL mapping using RNA-seq data

RNA-seq may replace gene expression microarrays in the near future. Using RNA-seq, the expression of a gene can be estimated using the total number of sequence reads mapped to that gene, known as the total read count (TReC). Traditional expression quantitative trait locus (eQTL) mapping methods, such as linear regression, can be applied to TReC measurements after they are properly normalized. In this article, we show that eQTL mapping, by directly modeling TReC using discrete distributions, has higher statistical power than the two-step approach: data normalization followed by linear regression. In addition, RNA-seq provides information on allele-specific expression (ASE) that is not available from microarrays. By combining the information from TReC and ASE, we can computationally distinguish cis- and trans-eQTL and further improve the power of cis-eQTL mapping. Both simulation and real data studies confirm the improved power of our new methods. We also discuss the design issues of RNA-seq experiments. Specifically, we show that by combining TReC and ASE measurements, it is possible to minimize cost and retain the statistical power of cis-eQTL mapping by reducing sample size while increasing the number of sequence reads per sample. In addition to RNA-seq data, our method can also be employed to study the genetic basis of other types of sequencing data, such as chromatin immunoprecipitation followed by DNA sequencing data. In this article, we focus on eQTL mapping of a single gene using the association-based method. However, our method establishes a statistical framework for future developments of eQTL mapping methods using RNA-seq data (e.g., linkage-based eQTL mapping), and the joint study of multiple genetic markers and/or multiple genes.     

An alternative approach for gene transfer in trees using wild-typeAgrobacterium strains

Micropropagated shoots of three forest tree species, poplar (Populus tremula × P.alba), wild cherry (Prunus avium L.) and walnut (Juglans nigra × J. regia), were inoculated each with six different wild-typeAgrobacterium strains. Poplar and wild cherry developed tumors that grew hormone-independently, whereas on walnut, gall formation was weak. On poplar and wild cherry, tumors induced by nopaline strains developed spontaneously shoots that had a normal phenotype and did not carry oncogenic T-DNA. From these observations, we have established a co-inoculation method to transform plants, using poplar as an experimental model. The method is based on inoculation of stem internodes with anAgrobacterium suspension containing both an oncogenic strain that induces shoot differentiation and a disarmed strain that provides the suitable genes in a binary vector. We used the vector pBI121 carryingneo (kanamycin resistance) anduidA (-glucuronidase) genes to facilitate early selection and screening. Poplar plants derived from kanamycin-resistant shoots that did not carry oncogenic T-DNA, were shown to contain and to expressneo anduidA genes. These results suggest that wild-typeAgrobacterium strains that induce shoot formation directly from tumors can be used as a general tool for gene transfer, avoiding difficult regeneration procedures.This work is dedicated to the late Marie-France Michel who initiated the poplar biotechnology project at INRA.     

Genetic mapping of a cloned sequence responsible for susceptibility to ecotropic murine leukemia viruses.

A mouse cDNA that confers susceptibility to ecotropic murine leukemia viruses following transfection into human EJ cells has been cloned and sequenced. We show that this sequence is likely to be Rec-1, the chromosome 5 locus originally defined by studies with somatic cell hybrids as responsible for virus susceptibility, and provide a specific chromosomal map position for this locus by analysis of an interspecies backcross. This locus maps in the distal region of chromosome 5 and is thus not within the cluster of retrovirus-related genes near the centromere.     

Prospects for using multimarker dihaploid lines for mapping the barley genome

Model population of spring barley Oregon Wolfe Pack dihaploid lines, derived from F1 of cross between dominant marker stock and recessive line, was studied on presence of morphological markers. For detection of DNA variability RAPD-analysis of these genotypes was carried out. Polymorphic fragment correlating with morphological trait was detected using primer P89.     

Congenic mapping and candidate gene analysis for streptozotocin-induced diabetes susceptibility locus on mouse chromosome 11

Streptozotocin (STZ) has been widely used to induce diabetes in rodents. Strain-dependent variation in susceptibility to STZ has been reported; however, the gene(s) responsible for STZ susceptibility has not been identified. Here, we utilized the A/J-11^SM consomic strain and a set of chromosome 11 (Chr. 11) congenic strains developed from A/J-11^SM to identify a candidate STZ-induced diabetes susceptibility gene. The A/J strain exhibited significantly higher susceptibility to STZ-induced diabetes than the A/J-11^SM strain, confirming the existence of a susceptibility locus on Chr. 11. We named this locus Stzds1 (STZ-induced diabetes susceptibility 1). Congenic mapping using the Chr. 11 congenic strains indicated that the Stzds1 locus was located between D11Mit163 (27.72 Mb) and D11Mit51 (36.39 Mb). The Mpg gene, which encodes N-methylpurine DNA glycosylase (MPG), a ubiquitous DNA repair enzyme responsible for the removal of alkylated base lesions in DNA, is located within the Stzds1 region. There is a close relationship between DNA alkylation at an early stage of STZ action and the function of MPG. A Sanger sequence analysis of the Mpg gene revealed five polymorphic sites in the A/J genome. One variant, p.Ala132Ser, was located in a highly conserved region among rodent species and in the minimal region for retained enzyme activity of MPG. It is likely that structural alteration of MPG caused by the p.Ala132Ser mutation elicits increased recognition and excision of alkylated base lesions in DNA by STZ.

     

Clustering by neurocognition for fine mapping of the schizophrenia susceptibility loci on chromosome 6p

Chromosome 6p is one of the most commonly implicated regions in the genome-wide linkage scans of schizophrenia, whereas further association studies for markers in this region were inconsistent likely due to heterogeneity. This study aimed to identify more homogeneous subgroups of families for fine mapping on regions around markers D6S296 and D6S309 (both in 6p24.3) as well as D6S274 (in 6p22.3) by means of similarity in neurocognitive functioning. A total of 160 families of patients with schizophrenia comprising at least two affected siblings who had data for eight neurocognitive test variables of the continuous performance test (CPT) and the Wisconsin card sorting test (WCST) were subjected to cluster analysis with data visualization using the test scores of both affected siblings. Family clusters derived were then used separately in family-based association tests for 64 single nucleotide polymorphisms (SNPs) covering the region of 6p24.3 and 6p22.3. Three clusters were derived from the family-based clustering, with deficit cluster 1 representing deficit on the CPT, deficit cluster 2 representing deficit on both the CPT and the WCST, and a third cluster of nondeficit. After adjustment using false discovery rate for multiple testing, SNP rs13873 and haplotype rs1225934-rs13873 on BMP6-TXNDC5 genes were significantly associated with schizophrenia for the deficit cluster 1 but not for the deficit cluster 2 or nondeficit cluster. Our results provide further evidence that the BMP6-TXNDC5 locus on 6p24.3 may play a role in the selective impairments on sustained attention of schizophrenia.