Cross-sample Normalization Research Articles

THU478 Unchecked Kinase Activation Underlies Pituitary Adenomas Causing Cushing’s Disease

Abstract Disclosure: D. Asuzu: None. D. Mullaney: None. Y. Li: None. D. Mandal: None. C. Tatsi: None. L.K. Nieman: None. P. Chittiboina: None. Introduction: Common mechanisms underlying canonically mutated (USP8, USP48 and BRAF) and ‘wildtype’ pituitary adenomas causing Cushing’s disease remain unknown. Using syngeneic, pairwise transcriptomic analysis at the single cell/nucleus resolution, we found a ubiquitous and robust overexpression of PPP1R17 in CD adenoma corticotrophs. The role of PPP1R17, an endogenous inhibitor of tumor-suppressor phosphatase PP2A in the embryonic context remains unknown in the pituitary. Here, we elucidated mechanistically, the effect of PPP1R17 on the phopshoproteome of CD adenomas. Methods: We quantitatively profiled the proteome and phosphoproteome of three pairs of CD adenomas and their syngeneic adjacent normal pituitary glands with TMT labeling and mass spectrometry (LC-MS/MS). We validated the mechanisms with murine corticotroph cells (mCort) overexpressing R17 (mCortR17) or GFP (mCortGFP). Elevated total and phosphprotein expression was assessed using R packages by performing cross sample normalization, noise filtering and differential protein abundance or phosphorylation analyses. Pathway enrichment analysis was performed on the KEGG database using ShinyGo 0.76.2. Results: We identified 37 proteins differentially phosphorylated in CD adenomas compared to adjacent normal pituitary glands, including the AP2-associated protein kinase 1 (Aak1) and the apoptotic chromatin condensation inducer Acin1. Of these 37 proteins, 30 (81%) were also differentially phosphorylated in mCortR17 versus mCortGFP cells, suggesting that differential phosphorylation in CD is driven by R17 hyperactivity. Of the 35 proteins overabundant in CD adenomas, 15 (43%) overlapped with mCortR17, including the Myc-associated protein (Max) and Presenilin 1 (Psen1). Proteins enriched in both human CD and mCort datasets were involved in NOTCH signaling and neurotransmitter release, while phospho-proteome profiling highlighted R17-mediated perturbation of RNA polymerase II transcription termination and mRNA processing. Conclusions:We first discovered a heavily dysregulated phosphoproteome in CD adenomas compared to syngeneic normal glands. We found that we could replicate the signature phosphoproteomic profile by overexpressing PPP1R17 in-vitro. These data indicate a central role for phosphatase (PP2A) inhibition in CD adenomas, and suggest potential novel anti-tumor targets for patients with CD. Presentation: Thursday, June 15, 2023

InterOpt: Improved gene expression quantification in qPCR experiments using weighted aggregation of reference genes

qPCR is still the gold standard for gene expression quantification. However, its accuracy is highly dependent on the normalization procedure. The conventional method involves using the geometric mean of multiple study-specific reference genes (RGs) expression for cross-sample normalization. While research on selecting stably expressed RGs is extensive, scant literature exists regarding the optimal approach for aggregating multiple RGs into a unified RG. In this paper, we introduce a family of scale-invariant functions as an alternative to the geometric mean aggregation. Our candidate method (weighted geometric mean minimizing standard deviation) demonstrated significantly better results compared to other proposed methods. We provide theoretical and experimental support for this finding using real data from solid tumors and liquid biopsies. Moreover, the closed form and regression-based solution enable efficient computation and straightforward adoption on various platforms. All the proposed methods have been implemented within an easy-to-use R package with graphics processing unit (GPU) acceleration.

CytoNorm: A Normalization Algorithm for Cytometry Data.

High‐dimensional flow cytometry has matured to a level that enables deep phenotyping of cellular systems at a clinical scale. The resulting high‐content data sets allow characterizing the human immune system at unprecedented single cell resolution. However, the results are highly dependent on sample preparation and measurements might drift over time. While various controls exist for assessment and improvement of data quality in a single sample, the challenges of cross‐sample normalization attempts have been limited to aligning marker distributions across subjects. These approaches, inspired by bulk genomics and proteomics assays, ignore the single‐cell nature of the data and risk the removal of biologically relevant signals. This work proposes CytoNorm, a normalization algorithm to ensure internal consistency between clinical samples based on shared controls across various study batches. Data from the shared controls is used to learn the appropriate transformations for each batch (e.g., each analysis day). Importantly, some sources of technical variation are strongly influenced by the amount of protein expressed on specific cell types, requiring several population‐specific transformations to normalize cells from a heterogeneous sample. To address this, our approach first identifies the overall cellular distribution using a clustering step, and calculates subset‐specific transformations on the control samples by computing their quantile distributions and aligning them with splines. These transformations are then applied to all other clinical samples in the batch to remove the batch‐specific variations. We evaluated the algorithm on a customized data set with two shared controls across batches. One control sample was used for calculation of the normalization transformations and the second control was used as a blinded test set and evaluated with Earth Mover's distance. Additional results are provided using two real‐world clinical data sets. Overall, our method compared favorably to standard normalization procedures. The algorithm is implemented in the R package “CytoNorm” and available via the following link: http://www.github.com/saeyslab/CytoNorm © 2019 The Authors. Cytometry Part A published by Wiley Periodicals, Inc. on behalf of International Society for Advancement of Cytometry.

Determination of qPCR Reference Genes Suitable for Normalizing Gene Expression in a Canine Model of Duchenne Muscular Dystrophy

Dogs with dystrophin-deficient muscular dystrophy are valuable models of the equivalent human disease, Duchenne Muscular Dystrophy (DMD): unlike the mdx mouse, these animals present a disease severity and progression that closely matches that found in human patients. Canine models are however less thoroughly characterised than the established mdx mouse in many aspects, including gene expression. Analysis of expression in muscle plays a key role in the study of DMD, allowing monitoring and assessment of disease progression, evaluation of novel biomarkers and gauging of therapeutic intervention efficacy. Appropriate normalization of expression data via carefully selected reference genes is consequently essential for accurate quantitative assessment. Unlike the expression profile of healthy skeletal muscle, the dystrophic muscle environment is highly dynamic: transcriptional profiles of dystrophic muscle might alter with age, disease progression, disease severity, genetic background and between muscle groups. The aim of this work was to identify reference genes suitable for normalizing gene expression in healthy and dystrophic dogs under various comparative scenarios. Using the delta-E50 MD canine model of DMD, we assessed a panel of candidate reference genes for stability of expression across healthy and dystrophic animals, at different ages and in different muscle groups. We show that the genes HPRT1, SDHA and RPL13a appear universally suitable for normalizing gene expression in healthy and dystrophic canine muscle, while other putative reference genes are exceptionally poor, and in the case of B2M, actively disease-correlated. Our findings suggest consistent cross-sample normalization is possible even throughout the dynamic progression of dystrophic pathology, and furthermore highlight the importance of empirical determination of suitable reference genes for neuromuscular diseases.

A Normalization-Free and Nonparametric Method Sharpens Large-Scale Transcriptome Analysis and Reveals Common Gene Alteration Patterns in Cancers.

Heterogeneity in transcriptional data hampers the identification of differentially expressed genes (DEGs) and understanding of cancer, essentially because current methods rely on cross-sample normalization and/or distribution assumption—both sensitive to heterogeneous values. Here, we developed a new method, Cross-Value Association Analysis (CVAA), which overcomes the limitation and is more robust to heterogeneous data than the other methods. Applying CVAA to a more complex pan-cancer dataset containing 5,540 transcriptomes discovered numerous new DEGs and many previously rarely explored pathways/processes; some of them were validated, both in vitro and in vivo, to be crucial in tumorigenesis, e.g., alcohol metabolism (ADH1B), chromosome remodeling (NCAPH) and complement system (Adipsin). Together, we present a sharper tool to navigate large-scale expression data and gain new mechanistic insights into tumorigenesis.

Abstract 4868: CODEX: a normalization and copy number variation detection method for whole-exome sequencing

Abstract Background: High throughput sequencing of DNA coding regions has become a common way of assaying genomic variation in the study of cancer. Copy number aberration (CNA) is an important type of genomic change, but detecting and characterizing CNA from whole-exome sequencing (WES) is challenging due to the high level of biases and artifacts. Methods: We propose CODEX, a normalization and CNA calling procedure for WES data. The Poisson latent factor model in CODEX includes terms that specifically remove biases due to GC content, exon capture and amplification efficiency, and latent systemic artifacts. CODEX also includes a Poisson likelihood-based recursive segmentation procedure that explicitly models the count-based WES data. CODEX can be used to detect both germline and somatic CNAs in cancer samples with or without matched normal. Results: Compared to existing approaches, CODEX is shown to be more effective in removing the biases in WES, and attains better sensitivity and specificity in detecting copy number aberrations by in silico spike-in studies. We further evaluate performance on 222 neuroblastoma samples with matched normal from the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) Project. We carry out systematic genome-wide analysis and detailed characterization of both germline and somatic copy number events. With a focus on a well-studied rare somatic CNV within the ATRX gene, we show that the cross-sample normalization procedure of CODEX is more effective in removing noise than the standard pipeline of normalizing the tumor against the matched normal, and that the segmentation procedure performs well in detecting CNVs with recurrent complex nested structures. For detecting germline mutations, CODEX is compared to existing methods on a population analysis of HapMap samples from the 1000 Genomes Project, and shown to be perform well on three microarray-based validation data sets. Conclusions: The cross-sample normalization procedure of CODEX, when applied to the matrix of tumor and normal samples, is more effective in reducing noise than normalizing each tumor to its matched normal. The somatic deletions in the ATRX region have a nested structure, which CODEX was able to recover. Through multiple types of validation, CODEX is shown to be applicable to a wide range of study designs for copy number estimation using WES data. Citation Format: Yuchao Jiang, Derek A. Oldridge, Sharon J. Diskin, Nancy R. Zhang. CODEX: a normalization and copy number variation detection method for whole-exome sequencing. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4868. doi:10.1158/1538-7445.AM2015-4868

Test set bias affects reproducibility of gene signatures.

Prior to applying genomic predictors to clinical samples, the genomic data must be properly normalized to ensure that the test set data are comparable to the data upon which the predictor was trained. The most effective normalization methods depend on data from multiple patients. From a biomedical perspective, this implies that predictions for a single patient may change depending on which other patient samples they are normalized with. This test set bias will occur when any cross-sample normalization is used before clinical prediction. We demonstrate that results from existing gene signatures which rely on normalizing test data may be irreproducible when the patient population changes composition or size using a set of curated, publicly available breast cancer microarray experiments. As an alternative, we examine the use of gene signatures that rely on ranks from the data and show why signatures using rank-based features can avoid test set bias while maintaining highly accurate classification, even across platforms. The code, data and instructions necessary to reproduce our entire analysis is available at https://github.com/prpatil/testsetbias.

CODEX: a normalization and copy number variation detection method for whole exome sequencing.

High-throughput sequencing of DNA coding regions has become a common way of assaying genomic variation in the study of human diseases. Copy number variation (CNV) is an important type of genomic variation, but detecting and characterizing CNV from exome sequencing is challenging due to the high level of biases and artifacts. We propose CODEX, a normalization and CNV calling procedure for whole exome sequencing data. The Poisson latent factor model in CODEX includes terms that specifically remove biases due to GC content, exon capture and amplification efficiency, and latent systemic artifacts. CODEX also includes a Poisson likelihood-based recursive segmentation procedure that explicitly models the count-based exome sequencing data. CODEX is compared to existing methods on a population analysis of HapMap samples from the 1000 Genomes Project, and shown to be more accurate on three microarray-based validation data sets. We further evaluate performance on 222 neuroblastoma samples with matched normals and focus on a well-studied rare somatic CNV within the ATRX gene. We show that the cross-sample normalization procedure of CODEX removes more noise than normalizing the tumor against the matched normal and that the segmentation procedure performs well in detecting CNVs with nested structures.

MRNA enrichment protocols determine the quantification characteristics of external RNA spike-in controls in RNA-Seq studies

RNA-Seq promises to be used in clinical settings as a gene-expression profiling tool; however, questions about its variability and biases remain and need to be addressed. Thus, RNA controls with known concentrations and sequence identities originally developed by the External RNA Control Consortium (ERCC) for microarray and qPCR platforms have recently been proposed for RNA-Seq platforms, but only with a limited number of samples. In this study, we report our analysis of RNA-Seq data from 92 ERCC controls spiked in a diverse collection of 447 RNA samples from eight ongoing studies involving five species (human, rat, mouse, chicken, and Schistosoma japonicum) and two mRNA enrichment protocols, i.e., poly(A) and RiboZero. The entire collection of datasets consisted of 15650143175 short sequence reads, 131603796 (i.e., 0.84%) of which were mapped to the 92 ERCC references. The overall ERCC mapping ratio of 0.84% is close to the expected value of 1.0% when assuming a 2.0% mRNA fraction in total RNA, but showed a difference of 2.8-fold across studies and 4.3-fold among samples from the same study with one tissue type. This level of fluctuation may prevent the ERCC controls from being used for cross-sample normalization in RNA-Seq. Furthermore, we observed striking biases of quantification between poly(A) and RiboZero which are transcript-specific. For example, ERCC-00116 showed a 7.3-fold under-enrichment in poly(A) compared to RiboZero. Extra care is needed in integrative analysis of multiple datasets and technical artifacts of protocol differences should not be taken as true biological findings.