Machine learning and data mining in complex genomic data—a review on the lessons learned in Genetic Analysis Workshop 19

In the analysis of current genomic data, application of machine learning and data mining techniques has become more attractive given the rising complexity of the projects. As part of the Genetic Analysis Workshop 19, approaches from this domain were explored, mostly motivated from two starting points. First, assuming an underlying structure in the genomic data, data mining might identify this and thus improve downstream association analyses. Second, computational methods for machine learning need to be developed further to efficiently deal with the current wealth of data.In the course of discussing results and experiences from the machine learning and data mining approaches, six common messages were extracted. These depict the current state of these approaches in the application to complex genomic data. Although some challenges remain for future studies, important forward steps were taken in the integration of different data types and the evaluation of the evidence. Mining the data for underlying genetic or phenotypic structure and using this information in subsequent analyses proved to be extremely helpful and is likely to become of even greater use with more complex data sets.     

Cell and tissue polarity in the intestinal tract during tumourigenesis: cells still know the right way up,but tissue organization is lost

Cell and tissue polarity are tightly coupled and are vital for normal tissue homeostasis. Changes in cellular and tissue organization are common to even early stages of disease, particularly cancer. The digestive tract is the site of the second most common cause of cancer deaths in the developed world. Tumours in this tissue arise in an epithelium that has a number of axes of cell and tissue polarity. Changes in cell and tissue polarity in response to genetic changes that are known to underpin disease progression provide clues about the link between molecular-, cellular- and tissue-based mechanisms that accompany cancer. Mutations in adenomatous polyposis coli (APC) are common to most colorectal cancers in humans and are sufficient to cause tumours in mouse intestine. Tissue organoids mimic many features of whole tissue and permit identifying changes at different times after inactivation of APC. Using gut organoids, we show that tissue polarity is lost very early during cancer progression, whereas cell polarity, at least apical–basal polarity, is maintained and changes only at later stages. These observations reflect the situation in tumours and validate tissue organoids as a useful system to investigate the relationship between cell polarity and tissue organization.     

rRNA-based profiling of bacteria in the axilla of healthy males suggests right-left asymmetry in bacterial activity

The activity of the human armpit microbiota triggers the formation of body odor. We used differential 16S rRNA gene (rDNA)- and rRNA-based terminal-restriction fragment length polymorphism fingerprinting in combination with cloning and sequencing to identify active members of the human armpit microbiota. DNA and RNA were isolated from skin scrub samples taken from both armpits of 10 preconditioned, healthy males. The fingerprint profiles indicated pronounced similarities between the armpit microbiota in the right and the left axillae of an individual test person, but larger differences between the axilla microbiota of different individuals. Using 16S rDNA and rRNA sequence data, the majority of peaks in the armpit profiles were assigned to bacteria affiliated with well-known genera of skin bacteria. The relative abundances of all groups were similar among the rDNA and rRNA samples, suggesting that all groups of armpit bacteria were active. Surprisingly, the relative abundance of sequences affiliated with Peptoniphilus sp. was by far and with statistical significance the highest in the rRNA samples of the right armpits. Thus, bacteria affiliated with Peptoniphilus sp. might have been particularly active in the right axillae of the test persons, possibly owing to the handedness of the test persons, which might cause different environmental conditions in the right axillae.     

Perkinsosis in the Manila clam Ruditapes philippinarum affects responses to the harmful-alga, Prorocentrum minimum

The dinoflagellate Prorocentrum minimum is increasingly recognized as a harmful algal bloom (HAB) species that affects filter-feeding shellfish. An experiment was done to investigate possible interactions between parasitic diseases and exposure to P. minimum in Manila clams, Ruditapes philippinarum. Manila clams, with variable levels of infection with Perkinsus olseni, were exposed for three or six days to the benign phytoplankton species Chaetoceros neogracile or a mixed diet of C. neogracile and P. minimum. After three or six days of exposure, clams were assessed individually for condition index, parasite status, and plasma and hemocyte parameters (morphological and functional) using flow-cytometry. Histological evaluation was also performed on individual clams to assess prevalence and intensity of parasitic infection, as well as other pathological conditions.Prorocentrum minimum caused several changes in Manila clams, especially after six days of exposure, such as decreased hemocyte phagocytosis and size and clam condition index. Pathological conditions observed in Manila clams exposed to P. minimum were hemocyte infiltration in the intestine and gonad follicles, myopathy, and necrosis of the intestine epithelial cells. The parasite P. olseni alone had no significant effect on Manila clams, nor did it modulate the hemocyte variables in clams exposed to P. minimum; however, the parasite did affect the pathological status of Manila clams exposed to the P. minimum culture, by causing atrophy and degeneration of residual ova in the gonadal follicles and hyaline degeneration of the muscle fibers, indicating synergistic effects of both stressors on the host over a short period of time. Additionally, an in vitro experiment also demonstrated detrimental effects of P. minimum and exudates upon P. olseni cells, thus suggesting HAB antagonistic suppression of transmission and proliferation of the parasite in the natural environment over a longer period of time. The results of this experiment demonstrate the complexity of interactions between host, parasite, and HAB.     

Single-Crossover Dynamics: Finite versus Infinite Populations

Populations evolving under the joint influence of recombination and resampling (traditionally known as genetic drift) are investigated. First, we summarize and adapt a deterministic approach, as valid for infinite populations, which assumes continuous time and single crossover events. The corresponding nonlinear system of differential equations permits a closed solution, both in terms of the type frequencies and via linkage disequilibria of all orders. To include stochastic effects, we then consider the corresponding finite-population model, the Moran model with single crossovers, and examine it both analytically and by means of simulations. Particular emphasis is on the connection with the deterministic solution. If there is only recombination and every pair of recombined offspring replaces their pair of parents (i.e., there is no resampling), then the expected type frequencies in the finite population, of arbitrary size, equal the type frequencies in the infinite population. If resampling is included, the stochastic process converges, in the infinite-population limit, to the deterministic dynamics, which turns out to be a good approximation already for populations of moderate size.     

Analysis of the transcriptional regulation of the FABP2 promoter haplotypes by PPARgamma/RXRalpha and Oct-1

Variants of the human intestinal fatty acid binding protein 2 gene (FABP2) are associated with traits of the metabolic syndrome. Relevant FABP2 promoter polymorphisms c.-80_-79insT, c.-136_-132delAGTAG, c.-168_-166delAAGinsT, c.-260G>A, c.-471G>A, and c.-778G>T result in two haplotypes A and B. Activation of haplotypes by rosiglitazone stimulated PPARgamma/RXRalpha leads to 2-fold higher activity of haplotype B than A. As shown by chimeric FABP2 promoter constructs, the higher responsiveness of FABP2 haplotype B is mainly but not solely determined by polymorphism c.-471G>A. As shown by EMSA and promoter-reporter assays, Oct-1 interacts with the -471 region of FABP2 promoters, induces the activities of both FABP2 promoter haplotypes and abolishes the different activities of haplotypes induced by rosiglitazone activated PPARgamma/RXRalpha. In conclusion, our findings suggest a functional role of PPARgamma/RXRalpha and Oct-1 in the regulation of the FABP2 gene.     

Adenomatous polyposis coli on microtubule plus ends in cell extensions can promote microtubule net growth with or without EB1

In interphase cells, the adenomatous polyposis coli (APC) protein accumulates on a small subset of microtubules (MTs) in cell protrusions, suggesting that APC may regulate the dynamics of these MTs. We comicroinjected a nonperturbing fluorescently labeled monoclonal antibody and labeled tubulin to simultaneously visualize dynamics of endogenous APC and MTs in living cells. MTs decorated with APC spent more time growing and had a decreased catastrophe frequency compared with non-APC-decorated MTs. Endogenous APC associated briefly with shortening MTs. To determine the relationship between APC and its binding partner EB1, we monitored EB1-green fluorescent protein and endogenous APC concomitantly in living cells. Only a small fraction of EB1 colocalized with APC at any one time. APC-deficient cells and EB1 small interfering RNA showed that EB1 and APC localized at MT ends independently. Depletion of EB1 did not change the growth-stabilizing effects of APC on MT plus ends. In addition, APC remained bound to MTs stabilized with low nocodazole, whereas EB1 did not. Thus, we demonstrate that the association of endogenous APC with MT ends correlates directly with their increased growth stability, that this can occur independently of its association with EB1, and that APC and EB1 can associate with MT plus ends by distinct mechanisms.     

Molecular signatures of cardiovascular disease risk: potential for test development and clinical application

Most cardiovascular diseases are multifactorial by etiology. As an example, the development of myocardial infarction is promoted by numerous risk factors, ranging from rather modifiable lifestyle habits (e.g. smoking, physical activity) to genetic predisposition. With respect to the latter, 15 years of candidate gene analyses have failed to explain the molecular basis for the genetic predisposition to myocardial infarction. By contrast, recent genome-wide association studies have identified chromosomal loci that reproducibly displayed some association with myocardial infarction risk. When molecular genetic studies of coronary artery disease were first begun, it was assumed that genetic factors would soon be routinely incorporated into risk prediction scores. A number of biomarkers have been identified and tested in combination with the classical risk factors for refined risk prediction. However, the strategy for individualized risk prediction by incorporation of new biomarkers in established scores has so far proven to be more difficult than at first hoped.