Standardizing workflows in imaging transcriptomics with the abagen toolbox

Gene expression fundamentally shapes the structural and functional architecture of the human brain. Open-access transcriptomic datasets like the Allen Human Brain Atlas provide an unprecedented ability to examine these mechanisms in vivo; however, a lack of standardization across research groups has given rise to myriad processing pipelines for using these data. Here, we develop the abagen toolbox, an open-access software package for working with transcriptomic data, and use it to examine how methodological variability influences the outcomes of research using the Allen Human Brain Atlas. Applying three prototypical analyses to the outputs of 750,000 unique processing pipelines, we find that choice of pipeline has a large impact on research findings, with parameters commonly varied in the literature influencing correlations between derived gene expression and other imaging phenotypes by as much as p ≥ 1:0. Our results further reveal an ordering of parameter importance, with processing steps that influence gene normalization yielding the greatest impact on downstream statistical inferences and conclusions. The presented work and the development of the abagen toolbox lay the foundation for more standardized and systematic research in imaging transcriptomics, and will help to advance future understanding of the influence of gene expression in the human brain.

Standardizing workflows in imaging transcriptomics with the abagen toolbox

Gene expression fundamentally shapes the structural and functional architecture of the human brain. Open-access transcriptomic datasets like the Allen Human Brain Atlas provide an unprecedented ability to examine these mechanisms in vivo; however, a lack of standardization across research groups has given rise to myriad processing pipelines for using these data. Here, we develop the abagen toolbox, an open-access software package for working with transcriptomic data, and use it to examine how methodological variability influences the outcomes of research using the Allen Human Brain Atlas. Applying three prototypical analyses to the outputs of 750,000 unique processing pipelines, we find that choice of pipeline has a large impact on research findings, with parameters commonly varied in the literature influencing correlations between derived gene expression and other imaging phenotypes by as much as ρ ≥ 1.0. Our results further reveal an ordering of parameter importance, with processing steps that influence gene normalization yielding the greatest impact on downstream statistical inferences and conclusions. The presented work and the development of the abagen toolbox lay the foundation for more standardized and systematic research in imaging transcriptomics, and will help to advance future understanding of the influence of gene expression in the human brain.

Using RNA-Seq to Identify Reference Genes of the Transition from Brown to White Adipose Tissue in Goats

Brown adipose tissues have unique non-shivering thermogenesis functions, can be found in newborn ruminate animals, and then are gradually replaced by white adipose tissues in adulthood. For the purpose of exploring the intrinsic mechanism underlying the conversion process from brown (BAT) to white adipose tissue (WAT), it is necessary to utilize Quantitative PCR (qPCR) to study gene expression profiling. In this study, we identified reference genes that were consistently expressed during the transformation from goat BAT to WAT using RNA-seq data. Then, twelve genes were evaluated as candidate reference genes for qPCR in goat perirenal adipose tissue using three tools (geNorm, Normfinder, and BestKeeper). In addition, the selected reference genes were used to normalize the gene expression of PGC-1α and GPAT4. It was found that traditional reference genes, such as GAPDH, RPLP0, HPRT1, and PPIA were not suitable for target gene normalization. In contrast, CTNNB, PFDN5, and EIF3M, selected from RNA sequencing data, showed the least variation and were recommended as the best reference genes during the transformation from BAT to WAT.

Systematic Normalization with Multiple Housekeeping Genes for the Discovery of Genetic Dependencies in Cancer

ABSTRACTCancer results from complex interactions between genes that are misregulated. Although our understanding of the contribution of single genes to cancer is expansive, the interplay between genes in the context of this devastating disease remains to be understood. Using the Genomic Data Commons Data Portal through National Cancer Institute, we randomly selected ten data sets of breast cancer gene expression, acquired by RNA sequencing to be subjected to a computational method for the exploration of genetic interactions at a large scale. We focused on genes that suppress genome instability (GIS genes) since function or expression of these genes is often altered in cancer.In this paper, we show how to discover pairs of genes whose expressions demonstrate patterns of correlation. To ensure an inter-comparison across data sets, we tested statistical normalization approaches derived from the expression of randomly selected single housekeeping genes, or from the average of three. In addition, we systematically selected ten housekeeping genes for the purpose of normalization. Using normalized expression data, we determined R2 values from linear models for all possible pairs of GIS genes and presented our results using heatmaps.Despite the heterogeneity of data, we observed that multiple gene normalization revealed more consistent correlations between pairs of genes, compared to using single gene expressions. We also noted that multiple gene normalization using ten genes outperformed normalization using three randomly selected genes. Since this study uses gene expression data from cancer tissues and begins to address the reproducibility of correlation between two genes, it complements other efforts to identify gene pairs that co-express in cancer cell lines. In the future, we plan to define consistent genetic correlations by using gene expression data derived from different types of cancer and multiple gene normalization.CCS CONCEPTSApplied computing → Computational biology.ACM Reference FormatOliver Bonham-Carter and Yee Mon Thu. 2019. Systematic Normalization with Multiple Housekeeping Genes for the Discovery of Genetic Dependencies in Cancer. In Niagara Falls, New York. ACM, New York, NY, USA, 10 pages. https://doi.org/10.1145/nnnnnnn.nnnnnnn

Identification of miR-328-3p as an endogenous reference gene for the normalization of miRNA expression data from patients with Diabetic Retinopathy

Diabetic Retinopathy, the main cause of visual loss and blindness among working population, is a complication of Diabetes mellitus (DM), which has been described as a major public health challenge, so it is important to identify biomarkers to predict and to stratify patient´s possibility for developing DR. MicroRNAs (miRNAs) are small non-coding RNA molecules that have showed to be promising disease biomarkers and association of miRNAs with the possibility to develop DR has been reported. However, evaluating miRNA expression involves normalization of RT-qPCR data using internal reference genes that should be properly determined, considering their impact on expression levels calculation and, until date, there is no unanimity on reference miRNAs for the investigation of circulating miRNAs in DR. We aimed to estimate the appropriateness of a group of miRNAs as normalizers to identify which might be considered steady internal reference genes in expression studies on DR plasma samples. Expression levels of candidates were analyzed in 60 healthy controls, 48 DM without DR patients and 62 DR patients with two statistical tools: NormFinder and RefFinder. MiR-328-3p was the most stable gene and we also investigated the effect of gene normalization, demonstrating that different normalization strategies have important implications for accurate data interpretation.

Comparison of Reliable Reference Genes Following Different Hormone Treatments by Various Algorithms for qRT-PCR Analysis of Metasequoia

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) is the most sensitive technique for evaluating gene expression levels. Choosing appropriate reference genes for normalizing target gene expression is important for verifying expression changes. Metasequoia is a high-quality and economically important wood species. However, few systematic studies have examined reference genes in Metasequoia. Here, the expression stability of 14 candidate reference genes in different tissues and following different hormone treatments were analyzed using six algorithms. Candidate reference genes were used to normalize the expression pattern of FLOWERING LOCUS T and pyrabactin resistance-like 8. Analysis using the GrayNorm algorithm showed that ACT2 (Actin 2), HIS (histone superfamily protein H3) and TATA (TATA binding protein) were stably expressed in different tissues. ACT2, EF1α (elongation factor-1 alpha) and HIS were optimal for leaves treated with the flowering induction hormone solution, while Cpn60β (60-kDa chaperonin β-subunit), GAPDH (glyceraldehyde-3-phosphate dehydrogenase) and HIS were the best reference genes for treated buds. EF1α, HIS and TATA were useful reference genes for accurate normalization in abscisic acid-response signaling. Our results emphasize the importance of validating reference genes for qRT-PCR analysis in Metasequoia. To avoid errors, suitable reference genes should be used for different tissues and hormone treatments to increase normalization accuracy. Our study provides a foundation for reference gene normalization when analyzing gene expression in Metasequoia.

GNormPlus: An Integrative Approach for Tagging Genes, Gene Families, and Protein Domains

The automatic recognition of gene names and their associated database identifiers from biomedical text has been widely studied in recent years, as these tasks play an important role in many downstream text-mining applications. Despite significant previous research, only a small number of tools are publicly available and these tools are typically restricted to detecting only mention level gene names or only document level gene identifiers. In this work, we report GNormPlus: an end-to-end and open source system that handles both gene mention and identifier detection. We created a new corpus of 694 PubMed articles to support our development of GNormPlus, containing manual annotations for not only gene names and their identifiers, but also closely related concepts useful for gene name disambiguation, such as gene families and protein domains. GNormPlus integrates several advanced text-mining techniques, including SimConcept for resolving composite gene names. As a result, GNormPlus compares favorably to other state-of-the-art methods when evaluated on two widely used public benchmarking datasets, achieving 86.7% F1-score on the BioCreative II Gene Normalization task dataset and 50.1% F1-score on the BioCreative III Gene Normalization task dataset. The GNormPlus source code and its annotated corpus are freely available, and the results of applying GNormPlus to the entire PubMed are freely accessible through our web-based tool PubTator.