Moritz Hess

Systematically evaluating interfaces for RNA-seq analysis from a life scientist perspective

RNA-sequencing (RNA-seq) has become an established way for measuring gene expression in model organisms and humans. While methods development for refining the corresponding data processing and analysis pipeline is ongoing, protocols for typical steps have been proposed and are widely used. Several user interfaces have been developed for making such analysis steps accessible to life scientists without extensive knowledge of command line tools. We performed a systematic search and evaluation of such interfaces to investigate to what extent these can indeed facilitate RNA-seq data analysis. We found a total of 29 open source interfaces, and six of the more widely used interfaces were evaluated in detail. Central criteria for evaluation were ease of configuration, documentation, usability, computational demand and reporting. No interface scored best in all of these criteria, indicating that the final choice will depend on the specific perspective of users and the corresponding weighting of criteria. Considerable technical hurdles had to be overcome in our evaluation. For many users, this will diminish potential benefits compared with command line tools, leaving room for future improvement of interfaces.

Partitioned learning of deep Boltzmann machines for SNP data

Motivation Learning the joint distributions of measurements, and in particular identification of an appropriate low‐dimensional manifold, has been found to be a powerful ingredient of deep leaning approaches. Yet, such approaches have hardly been applied to single nucleotide polymorphism (SNP) data, probably due to the high number of features typically exceeding the number of studied individuals. Results After a brief overview of how deep Boltzmann machines (DBMs), a deep learning approach, can be adapted to SNP data in principle, we specifically present a way to alleviate the dimensionality problem by partitioned learning. We propose a sparse regression approach to coarsely screen the joint distribution of SNPs, followed by training several DBMs on SNP partitions that were identified by the screening. Aggregate features representing SNP patterns and the corresponding SNPs are extracted from the DBMs by a combination of statistical tests and sparse regression. In simulated case‐control data, we show how this can uncover complex SNP patterns and augment results from univariate approaches, while maintaining type 1 error control. Time‐to‐event endpoints are considered in an application with acute myeloid leukemia patients, where SNP patterns are modeled after a pre‐screening based on gene expression data. The proposed approach identified three SNPs that seem to jointly influence survival in a validation dataset. This indicates the added value of jointly investigating SNPs compared to standard univariate analyses and makes partitioned learning of DBMs an interesting complementary approach when analyzing SNP data. Availability and implementation A Julia package is provided at ‘http://github.com/binderh/BoltzmannMachines.jl’. Contact binderh@imbi.uni‐freiburg.de Supplementary information Supplementary data are available at Bioinformatics online.

Suitability of Illumina deep mRNA sequencing for reliable gene expression profiling in a non-model conifer species ( Pseudotsuga menziesii )

Pseudotsuga menziesii (Douglas-fir) is an ideal model system to study the effect of local adaptation and intraspecific variation in transcriptome responses to the environment. Nonetheless, the lack of genomic resources and standardized microarray platforms for gene expression profiling has been a limitation to test the hypothesis on transcriptome organization and variation. Only recently, deep mRNA sequencing has become a promising alternative to overcome the present limitations. However, information on the transcript abundance distribution is needed for unbiased gene expression profiling from mRNA sequencing data. Since this information is not available for adult conifer needle tissue, we inferred the transcript abundance distribution and tested the effect of sequencing depth on the reliable detection and quantification of transcripts from the needle tissue of 50-year-old Douglas-fir trees. We obtained a similar distribution of GO-slim categories in our mRNA-sequencing libraries and in previously published putative unique transcripts (PUTs) for Douglas-fir, that were used as alignment reference. However, the GO-slim distribution in the Douglas-fir libraries and the Douglas-fir PUTs differed from the GO-slim distributions reported from mRNA deep sequencing libraries obtained from Arabidopsis thaliana leaf tissue. Apparently, several highly abundant PUTs associated with proteins involved in photosynthesis were limiting the benefits of increased sequencing depth. Simulations and empirical data indicated that a 3-fold increase from 5 to 15 million aligned reads results in about twice the number of PUTs that surpass the 100 aligned reads threshold that was used for robust transcript quantification.

Variation in short-term and long-term responses of photosynthesis and isoprenoid-mediated photoprotection to soil water availability in four Douglas-fir provenances

For long-lived forest tree species, the understanding of intraspecific variation among populations and their response to water availability can reveal their ability to cope with and adapt to climate change. Dissipation of excess excitation energy, mediated by photoprotective isoprenoids, is an important defense mechanism against drought and high light when photosynthesis is hampered. We used 50-year-old Douglas-fir trees of four provenances at two common garden experiments to characterize provenance-specific variation in photosynthesis and photoprotective mechanisms mediated by essential and non-essential isoprenoids in response to soil water availability and solar radiation. All provenances revealed uniform photoprotective responses to high solar radiation, including increased de-epoxidation of photoprotective xanthophyll cycle pigments and enhanced emission of volatile monoterpenes. In contrast, we observed differences between provenances in response to drought, where provenances sustaining higher CO2 assimilation rates also revealed increased water-use efficiency, carotenoid-chlorophyll ratios, pools of xanthophyll cycle pigments, β-carotene and stored monoterpenes. Our results demonstrate that local adaptation to contrasting habitats affected chlorophyll-carotenoid ratios, pool sizes of photoprotective xanthophylls, β-carotene, and stored volatile isoprenoids. We conclude that intraspecific variation in isoprenoid-mediated photoprotective mechanisms contributes to the adaptive potential of Douglas-fir provenances to climate change.

A deep learning approach for uncovering lung cancer immunome patterns

Tumor immune cell infiltration is a well known factor related to survival of cancer patients. This has led to deconvolution approaches that can quantify immune cell proportions for each individual. What is missing, is an approach for modeling joint patterns of different immune cell types. We adapt a deep learning approach, deep Boltzmann machines (DBMs), for modeling immune cell gene expression patterns in lung adenocarcinoma. Specifically, a partially partitioned training approach for dealing with a relatively large number of genes. We also propose a sampling-based approach that smooths the original data according to a trained DBM and can be used for visualization and clustering. The identified clusters can subsequently be judged with respect to association with clinical characteristics, such as tumor stage, providing an external criterion for selecting DBM network architecture and tuning parameters for training. We show that the hidden nodes of the trained networks cannot only be linked to clinical characteristics but also to specific genes, which are the visible nodes of the network. We find that hidden nodes that are linked to tumor stage and survival represent expression of T-cell and mast cell genes among others, probably reflecting specific immune cell infiltration patterns. Thus, DBMs, trained and selected by the proposed approach, might provide a useful tool for extracting immune cell gene expression patterns. In the case of lung adenocarcinomas, these patterns are linked to survival as well as other patient characteristics, which could be useful for uncovering the underlying biology.

Foliar nitrogen metabolism of adult Douglas-fir trees is affected by soil water availability and varies little among provenances

The coniferous forest tree Douglas-fir (Pseudotsuga menziesii) is native to the pacific North America, and is increasingly planted in temperate regions worldwide. Nitrogen (N) metabolism is of great importance for growth, resistance and resilience of trees. In the present study, foliar N metabolism of adult trees of three coastal and one interior provenance of Douglas-fir grown at two common gardens in southwestern Germany (Wiesloch, W; Schluchsee, S) were characterized in two subsequent years. Both the native North American habitats of the seed sources and the common garden sites in Germany differ in climate conditions. Total and mineral soil N as well as soil water content were higher in S compared to W. We hypothesized that i) provenances differ constitutively in N pool sizes and composition, ii) N pools are affected by environmental conditions, and iii) that effects of environmental factors on N pools differ among interior and coastal provenances. Soil water content strongly affected the concentrations of total N, soluble protein, total amino acids (TAA), arginine and glutamate. Foliar concentrations of total N, soluble protein, structural N and TAA of trees grown at W were much higher than in trees at S. Provenance effects were small but significant for total N and soluble protein content (interior provenance showed lowest concentrations), as well as arginine, asparagine and glutamate. Our data suggest that needle N status of adult Douglas-fir is independent from soil N availability and that low soil water availability induces a re-allocation of N from structural N to metabolic N pools. Small provenance effects on N pools suggest that local adaptation of Douglas-fir is not dominated by N conditions at the native habitats.

Cerebral ischemia-induced genes are increased in acute schizophrenia: An opportunity for clinical translation of genomic research findings

Schizophrenia is a brain disorder of unknown etiology. Brain imaging studies have revealed evidence for hypoperfusion of the frontal cortex (hypofrontality) and progressive brain volume reduction in schizophrenic patients. Mild cerebral ischemia (oligemia) has been postulated as a cause of the disorder. If the ischemia hypothesis for the adult brain is correct, genes induced by cerebral ischemia should be increased in the frontal cortex of schizophrenic patients during acute psychosis. Here, we show for the first time through a combined analysis of gene expression data from all the studies of the Stanley Brain Collection covering the Brodmann area 46 of the frontal cortex and employing the well-established Affymetrix HGU133a microarray platform that genes upregulated by cerebral ischemia are significantly overexpressed (4.5-fold) in the frontal cortex of acute schizophrenic patients (representation factor (RF) 4.5, p < 0.0002) and to a lesser degree in chronic patients (RF 3.9, p < 0.008) in comparison to normal controls. Neurodevelopmental-, repair-, inflammation- and synapse-related genes showed no significant change. The difference between acute and chronic schizophrenic patients regarding cerebral ischemia-induced genes was highly significant (RF 2.8, p < 0.00007). The results reported here are in line with evidence from biochemical, cellular, electroencephalographic, brain imaging, cerebral near-infrared spectroscopy, vascular, and genetic association studies. In summary, our genomic analysis revealed a clear ischemic signature in the frontal cortex of schizophrenia patients, confirming the prediction of the adult ischemia hypothesis for this disorder. This finding suggests new possibilities for the treatment and prevention of schizophrenia.