Feature selection in clinical proteomics: with great power comes great reproducibility

Exploiting the similarity of dissimilarities for biomedical applications and enhanced machine learning

The "similarity of dissimilarities" is an emerging paradigm in biomedical science with significant implications for protein function prediction, machine learning (ML), and personalized medicine. In protein function prediction, recognizing dissimilarities alongside similarities provides a more detailed understanding of evolutionary processes, allowing for a deeper exploration of regions that influence biological functionality. For ML models, incorporating dissimilarity measures helps avoid misleading results caused by highly correlated or similar data, addressing confounding issues like the Doppelgänger Effect. This leads to more accurate insights and a stronger understanding of complex biological systems. In the realm of personalized AI and precision medicine, the importance of dissimilarities is paramount. Personalized AI builds local models for each sample by identifying a network of neighboring samples. However, if the neighboring samples are too similar, it becomes difficult to identify factors critical to disease onset for the individual, limiting the effectiveness of personalized interventions or treatments. This paper discusses the "similarity of dissimilarities" concept, using protein function prediction, ML, and personalized AI as key examples. Integrating this approach into an analysis allows for the design of better, more meaningful experiments and the development of smarter validation methods, ensuring that the models learn in a meaningful way.

Assessing the precision of a detergent-assisted cartridge precipitation workflow for non-targeted quantitative proteomics

Detergent-based workflows incorporating sodium dodecyl sulfate (SDS) necessitate additional steps for detergent removal ahead of mass spectrometry (MS). These steps may lead to variable protein recovery, inconsistent enzyme digestion efficiency, and unreliable MS signals. To validate a detergent-based workflow for quantitative proteomics, we herein evaluate the precision of a bottom-up sample preparation strategy incorporating cartridge-based protein precipitation with organic solvent to deplete SDS. The variance of data-independent acquisition (SWATH-MS) data was isolated from sample preparation error by modelling the variance as a function of peptide signal intensity. Our SDS-assisted cartridge workflow yield a coefficient of variance (CV) of 13%-14%. By comparison, conventional (detergent-free) in-solution digestion increased the CV to 50%; in-gel digestion provided lower CVs between 14% and 20%. By filtering peptides predicting to display lower precision, we further enhance the validity of data in global comparative proteomics. These results demonstrate the detergent-based precipitation workflow is a reliable approach for in depth, label-free quantitative proteome analysis.

Integration of Omics Data and Network Models to Unveil Negative Aspects of SARS-CoV-2, from Pathogenic Mechanisms to Drug Repurposing

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) caused the COVID-19 health emergency, affecting and killing millions of people worldwide. Following SARS-CoV-2 infection, COVID-19 patients show a spectrum of symptoms ranging from asymptomatic to very severe manifestations. In particular, bronchial and pulmonary cells, involved at the initial stage, trigger a hyper-inflammation phase, damaging a wide range of organs, including the heart, brain, liver, intestine and kidney. Due to the urgent need for solutions to limit the virus' spread, most efforts were initially devoted to mapping outbreak trajectories and variant emergence, as well as to the rapid search for effective therapeutic strategies. Samples collected from hospitalized or dead COVID-19 patients from the early stages of pandemic have been analyzed over time, and to date they still represent an invaluable source of information to shed light on the molecular mechanisms underlying the organ/tissue damage, the knowledge of which could offer new opportunities for diagnostics and therapeutic designs. For these purposes, in combination with clinical data, omics profiles and network models play a key role providing a holistic view of the pathways, processes and functions most affected by viral infection. In fact, in addition to epidemiological purposes, networks are being increasingly adopted for the integration of multiomics data, and recently their use has expanded to the identification of drug targets or the repositioning of existing drugs. These topics will be covered here by exploring the landscape of SARS-CoV-2 survey-based studies using systems biology approaches derived from omics data, paying particular attention to those that have considered samples of human origin.

POSREG: proteomic signature discovered by simultaneously optimizing its reproducibility and generalizability

Mass spectrometry-based proteomic technique has become indispensable in current exploration of complex and dynamic biological processes. Instrument development has largely ensured the effective production of proteomic data, which necessitates commensurate advances in statistical framework to discover the optimal proteomic signature. Current framework mainly emphasizes the generalizability of the identified signature in predicting the independent data but neglects the reproducibility among signatures identified from independently repeated trials on different sub-dataset. These problems seriously restricted the wide application of the proteomic technique in molecular biology and other related directions. Thus, it is crucial to enable the generalizable and reproducible discovery of the proteomic signature with the subsequent indication of phenotype association. However, no such tool has been developed and available yet. Herein, an online tool, POSREG, was therefore constructed to identify the optimal signature for a set of proteomic data. It works by (i) identifying the proteomic signature of good reproducibility and aggregating them to ensemble feature ranking by ensemble learning, (ii) assessing the generalizability of ensemble feature ranking to acquire the optimal signature and (iii) indicating the phenotype association of discovered signature. POSREG is unique in its capacity of discovering the proteomic signature by simultaneously optimizing its reproducibility and generalizability. It is now accessible free of charge without any registration or login requirement at https://idrblab.org/posreg/.

Advances in Current Diabetes Proteomics: From the Perspectives of Label- free Quantification and Biomarker Selection

BACKGROUND

Due to its prevalence and negative impacts on both the economy and society, the diabetes mellitus (DM) has emerged as a worldwide concern. In light of this, the label-free quantification (LFQ) proteomics and diabetic marker selection methods have been applied to elucidate the underlying mechanisms associated with insulin resistance, explore novel protein biomarkers, and discover innovative therapeutic protein targets.

OBJECTIVE

The purpose of this manuscript is to review and analyze the recent computational advances and development of label-free quantification and diabetic marker selection in diabetes proteomics.

METHODS

Web of Science database, PubMed database and Google Scholar were utilized for searching label-free quantification, computational advances, feature selection and diabetes proteomics.

RESULTS

In this study, we systematically review the computational advances of label-free quantification and diabetic marker selection methods which were applied to get the understanding of DM pathological mechanisms. Firstly, different popular quantification measurements and proteomic quantification software tools which have been applied to the diabetes studies are comprehensively discussed. Secondly, a number of popular manipulation methods including transformation, pretreatment (centering, scaling, and normalization), missing value imputation methods and a variety of popular feature selection techniques applied to diabetes proteomic data are overviewed with objective evaluation on their advantages and disadvantages. Finally, the guidelines for the efficient use of the computationbased LFQ technology and feature selection methods in diabetes proteomics are proposed.

CONCLUSION

In summary, this review provides guidelines for researchers who will engage in proteomics biomarker discovery and by properly applying these proteomic computational advances, more reliable therapeutic targets will be found in the field of diabetes mellitus.

Why Inclusion Matters for Alzheimer's Disease Biomarker Discovery in Plasma

BACKGROUND

African American/Black adults have a disproportionate incidence of Alzheimer's disease (AD) and are underrepresented in biomarker discovery efforts.

OBJECTIVE

This study aimed to identify potential diagnostic biomarkers for AD using a combination of proteomics and machine learning approaches in a cohort that included African American/Black adults.

METHODS

We conducted a discovery-based plasma proteomics study on plasma samples (N = 113) obtained from clinically diagnosed AD and cognitively normal adults that were self-reported African American/Black or non-Hispanic White. Sets of differentially-expressed proteins were then classified using a support vector machine (SVM) to identify biomarker candidates.

RESULTS

In total, 740 proteins were identified of which, 25 differentially-expressed proteins in AD came from comparisons within a single racial and ethnic background group. Six proteins were differentially-expressed in AD regardless of racial and ethnic background. Supervised classification by SVM yielded an area under the curve (AUC) of 0.91 and accuracy of 86%for differentiating AD in samples from non-Hispanic White adults when trained with differentially-expressed proteins unique to that group. However, the same model yielded an AUC of 0.49 and accuracy of 47%for differentiating AD in samples from African American/Black adults. Other covariates such as age, APOE4 status, sex, and years of education were found to improve the model mostly in the samples from non-Hispanic White adults for classifying AD.

CONCLUSION

These results demonstrate the importance of study designs in AD biomarker discovery, which must include diverse racial and ethnic groups such as African American/Black adults to develop effective biomarkers.

Impact of the Identification Strategy on the Reproducibility of the DDA and DIA Results

Data-independent acquisition (DIA) is a promising technique for the proteomic analysis of complex protein samples. A number of studies have claimed that DIA experiments are more reproducible than data-dependent acquisition (DDA), but these claims are unsubstantiated since different data analysis methods are used in the two methods. Data analysis in most DIA workflows depends on spectral library searches, whereas DDA typically employs sequence database searches. In this study, we examined the reproducibility of the DIA and DDA results using both sequence database and spectral library search. The comparison was first performed using a cell lysate and then extended to an interactome study. Protein overlap among the technical replicates in both DDA and DIA experiments was 30% higher with library-based identifications than with sequence database identifications. The reproducibility of quantification was also improved with library search compared to database search, with the mean of the coefficient of variation decreasing more than 30% and a reduction in the number of missing values of more than 35%. Our results show that regardless of the acquisition method, higher identification and quantification reproducibility is observed when library search was used.

Quantitative proteomics to study aging in rabbit liver

Aging globally effects cellular and organismal metabolism across a range of mammalian species, including humans and rabbits. Rabbits (Oryctolagus cuniculus are an attractive model system of aging due to their genetic similarity with humans and their short lifespans. This model can be used to understand metabolic changes in aging especially in major organs such as liver where we detected pronounced variations in fat metabolism, mitochondrial dysfunction, and protein degradation. Such changes in the liver are consistent across several mammalian species however in rabbits the downstream effects of these changes have not yet been explored. We have applied proteomics to study changes in the liver proteins from young, middle, and old age rabbits using a multiplexing cPILOT strategy. This resulted in the identification of 2,586 liver proteins, among which 45 proteins had significant p < 0.05) changes with aging. Seven proteins were differentially-expressed at all ages and include fatty acid binding protein, aldehyde dehydrogenase, enoyl-CoA hydratase, 3-hydroxyacyl CoA dehydrogenase, apolipoprotein C3, peroxisomal sarcosine oxidase, adhesion G-protein coupled receptor, and glutamate ionotropic receptor kinate. Insights to how alterations in metabolism affect protein expression in liver have been gained and demonstrate the utility of rabbit as a model of aging.

Evaluating Combined Precursor Isotopic Labeling and Isobaric Tagging Performance on Orbitraps To Study the Peripheral Proteome of Alzheimer's Disease

Combined precursor isotopic labeling and isobaric tagging (cPILOT) is an enhanced multiplexing strategy currently capable of analyzing up to 24 samples simultaneously. This capability is especially helpful when studying multiple tissues and biological replicates in models of disease, such as Alzheimer's disease (AD). Here, cPILOT was used to study proteomes from heart, liver, and brain tissues in a late-stage amyloid precursor protein/presenilin-1 (APP/PS-1) human transgenic double-knock-in mouse model of AD. The original global cPILOT assay developed on an Orbitrap Velos instrument was transitioned to an Orbitrap Fusion Lumos instrument. The advantages of faster scan rates, lower limits of detection, and synchronous precursor selection on the Fusion Lumos afford greater numbers of isobarically tagged peptides to be quantified in comparison to the Orbitrap Velos. Parameters such as LC gradient, m/z isolation window, dynamic exclusion, targeted mass analyses, and synchronous precursor scan were optimized leading to >600 000 PSMs, corresponding to 6074 proteins. Overall, these studies inform of system-wide changes in brain, heart, and liver proteins from a mouse model of AD.

Advanced bioinformatics methods for practical applications in proteomics

Mass spectrometry (MS)-based proteomics has undergone rapid advancements in recent years, creating challenging problems for bioinformatics. We focus on four aspects where bioinformatics plays a crucial role (and proteomics is needed for clinical application): peptide-spectra matching (PSM) based on the new data-independent acquisition (DIA) paradigm, resolving missing proteins (MPs), dealing with biological and technical heterogeneity in data and statistical feature selection (SFS). DIA is a brute-force strategy that provides greater width and depth but, because it indiscriminately captures spectra such that signal from multiple peptides is mixed, getting good PSMs is difficult. We consider two strategies: simplification of DIA spectra to pseudo-data-dependent acquisition spectra or, alternatively, brute-force search of each DIA spectra against known reference libraries. The MP problem arises when proteins are never (or inconsistently) detected by MS. When observed in at least one sample, imputation methods can be used to guess the approximate protein expression level. If never observed at all, network/protein complex-based contextualization provides an independent prediction platform. Data heterogeneity is a difficult problem with two dimensions: technical (batch effects), which should be removed, and biological (including demography and disease subpopulations), which should be retained. Simple normalization is seldom sufficient, while batch effect-correction algorithms may create errors. Batch effect-resistant normalization methods are a viable alternative. Finally, SFS is vital for practical applications. While many methods exist, there is no best method, and both upstream (e.g. normalization) and downstream processing (e.g. multiple-testing correction) are performance confounders. We also discuss signal detection when class effects are weak.

Reproducibility of biomarker identifications from mass spectrometry proteomic data in cancer studies

Reproducibility of disease signatures and clinical biomarkers in multi-omics disease analysis has been a key challenge due to a multitude of factors. The heterogeneity of the limited sample, various biological factors such as environmental confounders, and the inherent experimental and technical noises, compounded with the inadequacy of statistical tools, can lead to the misinterpretation of results, and subsequently very different biology. In this paper, we investigate the biomarker reproducibility issues, potentially caused by differences of statistical methods with varied distribution assumptions or marker selection criteria using Mass Spectrometry proteomic ovarian tumor data. We examine the relationship between effect sizes, p values, Cauchy p values, False Discovery Rate p values, and the rank fractions of identified proteins out of thousands in the limited heterogeneous sample. We compared the markers identified from statistical single features selection approaches with machine learning wrapper methods. The results reveal marked differences when selecting the protein markers from varied methods with potential selection biases and false discoveries, which may be due to the small effects, different distribution assumptions, and p value type criteria versus prediction accuracies. The alternative solutions and other related issues are discussed in supporting the reproducibility of findings for clinical actionable outcomes.

Deeper investigation into the utility of functional class scoring in missing protein prediction from proteomics data

Functional Class Scoring (FCS) is a network-based approach previously demonstrated to be powerful in missing protein prediction (MPP). We update its performance evaluation using data derived from new proteomics technology (SWATH) and also checked for reproducibility using two independent datasets profiling kidney tissue proteome. We also evaluated the objectivity of the FCS p-value, and followed up on the value of MPP from predicted complexes. Our results suggest that (1) FCS [Formula: see text]-values are non-objective, and are confounded strongly by complex size, (2) best recovery performance do not necessarily lie at standard [Formula: see text]-value cutoffs, (3) while predicted complexes may be used for augmenting MPP, they are inferior to real complexes, and are further confounded by issues relating to network coverage and quality and (4) moderate sized complexes of size 5 to 10 still exhibit considerable instability, we find that FCS works best with big complexes. While FCS is a powerful approach, blind reliance on its non-objective [Formula: see text]-value is ill-advised.

Urinary Biomarkers of Renal Fibrosis

Renal fibrosis is the common pathological pathway of progressive CKD. The commonly used biomarkers in clinical practice are not optimal to detect injury or predict prognosis. Therefore, it is crucial to develop novel biomarkers to allow prompt intervention. Urine serves as a valuable resource of biomarker discovery for kidney diseases. Owing to the rapid development of omics platforms and bioinformatics, research on novel urinary biomarkers for renal fibrosis has proliferated in recent years. In this chapter, we discuss the current status and provide basic knowledge in this field. We present novel promising biomarkers including tubular injury markers, proteins related to activated inflammation/fibrosis pathways, CKD273, transcriptomic biomarkers, as well as metabolomic biomarkers. Furthermore, considering the complex nature of the pathogenesis of renal fibrosis, we also highlight the combination of biomarkers to further improve the diagnostic and prognostic performance.

Turning straw into gold: building robustness into gene signature inference

Reproducible and generalizable gene signatures are essential for clinical deployment, but are hard to come by. The primary issue is insufficient mitigation of confounders: ensuring that hypotheses are appropriate, test statistics and null distributions are appropriate, and so on. To further improve robustness, additional good analytical practices (GAPs) are needed, namely: leveraging existing data and knowledge; careful and systematic evaluation of gene sets, even if they overlap with known sources of confounding; and rigorous testing of inferred signatures against as many published data sets as possible. Here, using a re-examination of a breast cancer data set and 48 published signatures, we illustrate the value of adopting these GAPs.

Why breast cancer signatures are no better than random signatures explained

Random signature superiority (RSS) occurs when random gene signatures outperform published and/or known signatures. Unlike reproducibility and generalizability issues, RSS is relatively underexplored. Yet, understanding it is imperative for better analytical outcome. In breast cancer, RSS correlates strongly with enrichment for proliferation genes and signature size. Removal of proliferation genes from random signatures reduces the predictive power of random signatures. Almost all genes are correlated to a certain extent with the proliferation signature, making complete elimination of its confounding effects impossible. RSS goes beyond breast cancer, because it also exists in other diseases; it is especially strong in other cancers in a platform-independent manner, and less severe, but present nonetheless, in nonproliferative diseases.

A Review of Matched-pairs Feature Selection Methods for Gene Expression Data Analysis

With the rapid accumulation of gene expression data from various technologies, e.g., microarray, RNA-sequencing (RNA-seq), and single-cell RNA-seq, it is necessary to carry out dimensional reduction and feature (signature genes) selection in support of making sense out of such high dimensional data. These computational methods significantly facilitate further data analysis and interpretation, such as gene function enrichment analysis, cancer biomarker detection, and drug targeting identification in precision medicine. Although numerous methods have been developed for feature selection in bioinformatics, it is still a challenge to choose the appropriate methods for a specific problem and seek for the most reasonable ranking features. Meanwhile, the paired gene expression data under matched case-control design (MCCD) is becoming increasingly popular, which has often been used in multi-omics integration studies and may increase feature selection efficiency by offsetting similar distributions of confounding features. The appropriate feature selection methods specifically designed for the paired data, which is named as matched-pairs feature selection (MPFS), however, have not been maturely developed in parallel. In this review, we compare the performance of 10 feature-selection methods (eight MPFS methods and two traditional unpaired methods) on two real datasets by applied three classification methods, and analyze the algorithm complexity of these methods through the running of their programs. This review aims to induce and comprehensively present the MPFS in such a way that readers can easily understand its characteristics and get a clue in selecting the appropriate methods for their analyses.

Dealing with Confounders in Omics Analysis

The Anna Karenina effect is a manifestation of the theory-practice gap that exists when theoretical statistics are applied on real-world data. In the course of analyzing biological data for differential features such as genes or proteins, it derives from the situation where the null hypothesis is rejected for extraneous reasons (or confounders), rather than because the alternative hypothesis is relevant to the disease phenotype. The mechanics of applying statistical tests therefore must address and resolve confounders. It is inadequate to simply rely on manipulating the P-value. We discuss three mechanistic elements (hypothesis statement construction, null distribution appropriateness, and test-statistic construction) and suggest how they can be designed to foil the Anna Karenina effect to select phenotypically relevant biological features.

Can Peripheral Blood-Derived Gene Expressions Characterize Individuals at Ultra-high Risk for Psychosis?

The ultra-high risk (UHR) state was originally conceived to identify individuals at imminent risk of developing psychosis. Although recent studies have suggested that most individuals designated UHR do not, they constitute a distinctive group, exhibiting cognitive and functional impairments alongside multiple psychiatric morbidities. UHR characterization using molecular markers may improve understanding, provide novel insight into pathophysiology, and perhaps improve psychosis prediction reliability. Whole-blood gene expressions from 56 UHR subjects and 28 healthy controls are checked for existence of a consistent gene expression profile (signature) underlying UHR, across a variety of normalization and heterogeneity-removal techniques, including simple log-conversion, quantile normalization, gene fuzzy scoring (GFS), and surrogate variable analysis. During functional analysis, consistent and reproducible identification of important genes depends largely on how data are normalized. Normalization techniques that address sample heterogeneity are superior. The best performer, the unsupervised GFS, produced a strong and concise 12-gene signature, enriched for psychosis-associated genes. Importantly, when applied on random subsets of data, classifiers built with GFS are "meaningful" in the sense that the classifier models built using genes selected after other forms of normalization do not outperform random ones, but GFS-derived classifiers do. Data normalization can present highly disparate interpretations on biological data. Comparative analysis has shown that GFS is efficient at preserving signals while eliminating noise. Using this, we demonstrate confidently that the UHR designation is well correlated with a distinct blood-based gene signature.

Transcriptome Analysis Reveals Neuroprotective aspects of Human Reactive Astrocytes induced by Interleukin 1β

Reactive astrogliosis is a critical process in neuropathological conditions and neurotrauma. Although it has been suggested that it confers neuroprotective effects, the exact genomic mechanism has not been explored. The prevailing dogma of the role of astrogliosis in inhibition of axonal regeneration has been challenged by recent findings in rodent model’s spinal cord injury, demonstrating its neuroprotection and axonal regeneration properties. We examined whether their neuroprotective and axonal regeneration potentials can be identify in human spinal cord reactive astrocytes in vitro. Here, reactive astrogliosis was induced with IL1β. Within 24 hours of IL1β induction, astrocytes acquired reactive characteristics. Transcriptome analysis of over 40000 transcripts of genes and analysis with PFSnet subnetwork revealed upregulation of chemokines and axonal permissive factors including FGF2, BDNF, and NGF. In addition, most genes regulating axonal inhibitory molecules, including ROBO1 and ROBO2 were downregulated. There was no increase in the gene expression of “Chondroitin Sulfate Proteoglycans” (CSPGs’) clusters. This suggests that reactive astrocytes may not be the main CSPG contributory factor in glial scar. PFSnet analysis also indicated an upregulation of “Axonal Guidance Signaling” pathway. Our result suggests that human spinal cord reactive astrocytes is potentially neuroprotective at an early onset of reactive astrogliosis.

NetProt: Complex-based Feature Selection

Protein complex-based feature selection (PCBFS) provides unparalleled reproducibility with high phenotypic relevance on proteomics data. Currently, there are five PCBFS paradigms, but not all representative methods have been implemented or made readily available. To allow general users to take advantage of these methods, we developed the R-package NetProt, which provides implementations of representative feature-selection methods. NetProt also provides methods for generating simulated differential data and generating pseudocomplexes for complex-based performance benchmarking. The NetProt open source R package is available for download from https://github.com/gohwils/NetProt/releases/ , and online documentation is available at http://rpubs.com/gohwils/204259 .

Integration of Proteomics and Metabolomics in Exploring Genetic and Rare Metabolic Diseases

BACKGROUND

Inherited metabolic disorders or inborn errors of metabolism are caused by deficiency of enzymatic activities in the catabolism of amino acids, carbohydrates, or lipids. These disorders include aminoacidopathies, urea cycle defects, organic acidemias, defects of oxidation of fatty acids, and lysosomal storage diseases. Inborn errors of metabolism constitute a significant proportion of genetic diseases, particularly in children; however, they are individually rare. Clinical phenotypes are very variable, some of them remain asymptomatic, others manifest metabolic decompensation in neonatal age, and others encompass mental retardation at later age. The clinical manifestation of these disorders can involve different organs and/or systems. Some disorders are easily managed if promptly diagnosed and treated, whereas in other cases neither diet, vitamin therapy, nor transplantation appears to prevent multi-organ impairment.

SUMMARY

Here, we discuss the principal challenges of metabolomics and proteomics in inherited metabolic disorders. We review the recent developments in mass spectrometry-based proteomic and metabolomic strategies. Mass spectrometry has become the most widely used platform in proteomics and metabolomics because of its ability to analyze a wide range of molecules, its optimal dynamic range, and great sensitivity. The fast measurement of a broad spectrum of metabolites in various body fluids, also collected in small samples like dried blood spots, have been facilitated by the use of mass spectrometry-based techniques. These approaches have enabled the timely diagnosis of inherited metabolic disorders, thereby facilitating early therapeutic intervention. Due to its analytical features, proteomics is suited for the basic investigation of inborn errors of metabolism. Modern approaches enable detailed functional characterization of the pathogenic biochemical processes, as achieved by quantification of proteins and identification of their regulatory chemical modifications.

KEY MESSAGE

Mass spectrometry-based "omics" approaches most frequently used to study the molecular mechanisms underlying inherited metabolic disorders pathophysiology are described.

Class-paired Fuzzy SubNETs: A paired variant of the rank-based network analysis family for feature selection based on protein complexes

Identifying reproducible yet relevant protein features in proteomics data is a major challenge. Analysis at the level of protein complexes can resolve this issue and we have developed a suite of feature-selection methods collectively referred to as Rank-Based Network Analysis (RBNA). RBNAs differ in their individual statistical test setup but are similar in the sense that they deploy rank-defined weights among proteins per sample. This procedure is known as gene fuzzy scoring. Currently, no RBNA exists for paired-sample scenarios where both control and test tissues originate from the same source (e.g. same patient). It is expected that paired tests, when used appropriately, are more powerful than approaches intended for unpaired samples. We report that the class-paired RBNA, PPFSNET, dominates in both simulated and real data scenarios. Moreover, for the first time, we explicitly incorporate batch-effect resistance as an additional evaluation criterion for feature-selection approaches. Batch effects are class irrelevant variations arising from different handlers or processing times, and can obfuscate analysis. We demonstrate that PPFSNET and an earlier RBNA, PFSNET, are particularly resistant against batch effects, and only select features strongly correlated with class but not batch.