Biology subjects

Asiaee, A.

Publications and source records attributed to Asiaee, A..

11 recordsLinked to original sources

bioRxiv2026

Context-dependent correlations mislead transcriptomic network inference in bulk and single-cell data

BackgroundCorrelation is the dominant input to co-expression module discovery and miRNA-target inference. Both rely on an implicit assumption: a Pearson coefficient pooled across heterogeneous samples, whether tissues, cancer types, or cell types, estimates one biologically meaningful quantity. Simpsons paradox makes this assumption fragile in principle, since between- group mean shifts can dominate or reverse within-group associations. How often this happens in real transcriptomic data has not been quantified. ResultsAcross 8,890 TCGA tumors from 31 cancer cohorts and 23,170,038 miRNA-mRNA pairs, 94.8% of pairs showed both positive and negative within-cohort correlations. Restricting to the high-variance domain of one million pairs, 13.3% of pooled correlations with |rglobal|[≥]0.2 reversed against the within-cohort majority at sign tolerance{varepsilon} = 0.05. Heterogeneity was the rule rather than the exception (median I2 = 0.86, IQR 0.80-0.90), and 99.5% of pairs rejected equal correlation across cohorts at FDR < 0.05. Of 692,770 experimentally validated miRTarBase v10 targets measurable in our data, only 0.9% were uniformly negative across cohorts. The pattern recurred across modalities. In GTEx, 21.0% of pooled signs disagreed with the tissue majority, and 23.5% of pairs flipped sign after tissue-mean removal. In 10x PBMC scRNA-seq, 13.1% of gene-gene correlations flipped after cell-type-mean removal; in CITE-seq, 37.9% of protein-RNA pairs flipped under a joint WNN partition of cells. Refining context reduced reversal, though by how much depended on the partition: within BRCA, 5.5% of pairs reversed under molecular PAM50 subtypes versus 0.35% under clinical IHC receptor status, and refining T cells into transcriptome-defined subtypes cut PBMC reversal from 11.8% to 0.13%. ConclusionsA single pooled correlation coefficient can invert direction relative to its within-context constituents at rates that are not negligible. Correlations should be reported with their context: the within-context distribution, a heterogeneity statistic, and a diagnostic that separates between-context mean shifts from within-context association. We provide a small R interface that computes these summaries.

bioinformatics↗

bioRxiv2026

Tissue-aware elastic net decomposition reveals shared and lineage-specific drug response biomarkers

MotivationComputational models that predict cancer drug response from genomic features are central to biomarker discovery, yet a recent audit found data leakage in 72% of 32 published methods, and complex models offer little interpretability while only modestly exceeding simple baselines under honest evaluation. Tissue lineage is a largely untapped source of legitimate inductive bias, but existing tissue-aware methods neither separate pan-cancer from lineage-specific signal nor report leakage-free performance. ResultsWe introduce the Data Shared Elastic Net (DSEN), a tissue-aware regression that decomposes each drugs model into a shared coefficient block common to all lineages and tissue-specific deviation blocks. Under leakage-free cross-validation across 265 drugs, 1,462 cell lines and 31 tissue lineages, DSEN improved mean squared error over a standard elastic net for 92.5% of drugs (mean 4.95%) while selecting 58% fewer stable shared features. Shared coefficients generalized to held-out tissues (59% tissue-level win rate) and recurrently recovered transferable pathway modules (p53, MAPK), whereas tissue blocks captured lineage markers such as the skin MITF /S100B program. The closest tissue-aware comparator, TG-LASSO, performed worse than the tissue-agnostic baseline (-13.8% mean MSE). Ablation shows tissue-aware modeling helps most when features are scarce, with no single modality dominating. Availability and implementationhttps://github.com/AsiaeeLab/tissue-aware-drug-response. Contactamir.asiaeetaheri@vumc.org

bioinformatics↗

bioRxiv2026

A shape-constrained regression and wild bootstrap framework for reproducible drug synergy testing

High-throughput drug combination screens require methods to identify synergistic pairs, yet widely used synergy scores lack statistical inference and can fail when parametric dose-response fits do not converge. We present SIR (Synergy via Isotonic Regression), a nonparametric framework that defines interaction as deviation from a monotone-additive null, fit by 2D isotonic regression. A degrees-of-freedom-corrected wild bootstrap yields calibrated p-values for each dose-response matrix. On DrugCombDB, SIR interaction surfaces achieve higher replicate concordance (median correlation 0.91 across 1,839 replicate pairs) than all baselines (0.53-0.74), while avoiding Loewes 20.9% and ZIPs 3.6% failure rates. The fitted surface also predicts missing wells (median holdout RMSE 0.040). By replacing heuristic scores with calibrated effect sizes and p-values, SIR enables principled hit calling and error-rate control in large screens.

bioinformatics↗

bioRxiv2026

Widespread data leakage inflates performance estimates in cancer drug response prediction

Drug response prediction models guide biomarker discovery, yet their reported accuracy depends on rigorous cross-validation (CV). We show that supervised feature screening applied to all samples before CV, a widespread practice, introduces data leakage that systematically underestimates prediction error. Across 265 drugs and 1,462 cancer cell lines, leakage-free CV raises mean squared error by 16.6% on average, with near-zero feature overlap between leaked and corrected pipelines (mean Jaccard 0.18; 36.2% of drugs share no features). Despite selecting five times more features, the leaked pipeline recovers known drug targets at nearly the same rate as the corrected pipeline, indicating that the inflated feature sets capture statistical artifacts rather than biological signal. A code-level audit of 32 published methods (2017-2024) confirms leakage in 23 (72%), spanning five distinct modes and cited over 3,000 times. The magnitude of inflation from this single leakage mode is comparable to improvement margins typically claimed over elastic net baselines, raising the possibility that some reported advances reflect evaluation artifacts. We provide a leakage taxonomy, audit guide and reference implementation for leakage-free evaluation.

bioinformatics↗

bioRxiv2025

Co-evolved Partners of Immunity: A Trait-Based Map of Human Keystone Organisms

Persistent human-adapted microbes can act as immunological "keystones," organizing host defense across tissues and shaping vulnerability under immune perturbation. More generally, tissue immunity is calibrated by persistent niche-resident organisms that tune compartment-specific thresholds of cytotoxicity and peripheral tolerance; keystone organisms represent the apex subset with multi-niche scope. Here we operationalize keystone organisms as pathogens whose containment requires coordinated engagement of multiple immune arms and whose residence is structured across anatomical niches. Using 18 curated immunological and evolutionary traits across 43 organisms, unsupervised analyses resolved four reproducible archetypes and identified a compact keystone set dominated by persistent herpesviruses and Mycobacterium tuberculosis. We then translated the clinical literature into a pathogenximmune-perturbationxniche tensor capturing where and when each organism emerges under defined immune deficits. We quantified "diagnostic breadth" with two complementary summaries: immune breadth (diversity of perturbations associated with emergence) and niche breadth (diversity of anatomical sites). Clinical emergence patterns perfectly separated trait-defined keystones from all other organisms and highlighted expanded niche breadth as the primary discriminator, whereas immune breadth showed no significant group separation. Finally, a mechanistic model integrating barrier disruption, latent reservoir activation, and tissue-resident immune control predicted clinical emergence from first principles--without fitting parameters to individual pathogens--and ranked true emergences 2.9-fold above chance among its highest-confidence predictions. Together, these results link evolutionary adaptation to clinically readable patterns of reactivation, motivate archetype-aware surveillance under immunosuppression, and provide a framework for immunogen design that prioritizes conserved, functionally constrained targets. Because the clinical tensor is literature-curated and sparse, "perfect separation" refers to keystone-vs-other discrimination within this dataset and is not a claim of universal out-of-sample performance.

immunology↗

bioRxiv2025

AUC-PR is a More Informative Metric for Assessing the Biological Relevance of In Silico Cellular Perturbation Prediction Models

In silico perturbation models, computational methods which can predict cellular responses to perturbations, present an opportunity to reduce the need for costly and time-intensive in vitro experiments. Many recently proposed models predict high-dimensional cellular responses, such as gene or protein expression to perturbations such as gene knockout or drugs. However, evaluating in silico performance has largely relied on metrics such as R2, which assess overall prediction accuracy but fail to capture biologically significant outcomes like the identification of differentially expressed genes. In this study, we present a novel evaluation framework that introduces the AUC-PR metric to assess the precision and recall of DE gene predictions. By applying this framework to both single-cell and pseudo-bulked datasets, we systematically benchmark simple and advanced computational models. Our results highlight a significant discrepancy between R2 and AUC-PR, with models achieving high R2 values but struggling to identify Differentially expressed genes accurately, as reflected in their low AUC-PR values. This finding underscores the limitations of traditional evaluation metrics and the importance of biologically relevant assessments. Our framework provides a more comprehensive understanding of model capabilities, advancing the application of computational approaches in cellular perturbation research.

bioinformatics↗

bioRxiv2024

Quantifying Differential Rhythmicity based on Effect Sizes with LimoRhyde2

Current methods for assessing differential rhythmicity in genomic data focus on hypothesis testing and model selection, often assuming sinusoidal rhythms. A more appropriate approach is to estimate differences in rhythmic properties between two or more conditions using effect sizes. To address this gap, we extend LimoRhyde2, a method for quantifying rhythm-related effect sizes and their uncertainty in genome-scale data, to enable differential rhythmicity analyses. Through extensive testing, we validate the method for differential rhythmicity analysis and showcase how it improves biological interpretation for circadian systems biology.

bioinformatics↗

bioRxiv2024

Improving Drug Sensitivity Prediction and Inference by Multitask Learning

The development of models to predict sensitivity to anticancer drugs is an area of significant interest, given the diverse responses to treatment among patients and the considerable expense and time involved in anticancer drug development. Leveraging "omic" data and anticancer response information from the Cancer Cell Line Encyclopedia, we propose a novel approach utilizing multitask learning to enhance prediction accuracy and inference. We extended a multitask learning framework called the Data Shared Lasso to develop the Data Shared Elastic Net. This enabled the construction of tissue-specific models with information sharing while maintaining the attractive properties of Elastic Net regression. By employing this approach, we observed improvements in prediction accuracy compared to single-task Elastic Net models, particularly for cell lines displaying high sensitivity to treatment. Furthermore, the Data Shared Elastic Net facilitated the identification of predictors for anticancer drug sensitivity within specific tissue types, shedding light on cellular pathways targeted by these drugs across tissues. We also investigated the impact of data leakage on modeling outcomes from previous studies, which led to underestimating prediction error and erroneous inferences

bioinformatics↗

bioRxiv2023

Transcriptome Complexity Disentangled: A Regulatory Elements Approach

AbstractTranscription factors (TFs) and microRNAs (miRNAs) are fundamental regulators of gene expression, cell state, and biological processes. This study investigated whether a small subset of TFs and miRNAs could accurately predict genome-wide gene expression. We analyzed 8895 samples across 31 cancer types from The Cancer Genome Atlas and identified 28 miRNA and 28 TF clusters using unsupervised learning. Medoids of these clusters could differentiate tissues of origin with 92.8% accuracy, demonstrating their biological relevance. We developed Tissue-Agnostic and Tissue-Aware models to predict 20,000 gene expressions using the 56 selected medoid miRNAs and TFs. The Tissue- Aware model attained an R2 of 0.70 by incorporating tissue-specific information. Despite measuring only 1/400th of the transcriptome, the prediction accuracy was comparable to that achieved by the 1000 landmark genes. This suggests the transcriptome has an intrinsically low-dimensional structure that can be captured by a few regulatory molecules. Our approach could enable cheaper transcriptome assays and analysis of low-quality samples. It also provides insights into genes that are heavily regulated by miRNAs/TFs versus alternative mechanisms. However, model transportability was impacted by dataset discrepancies, especially in miRNA distribution. Overall, this study demonstrates the potential of a biology-guided approach for robust transcriptome representation.

bioinformatics↗

bioRxiv2020

Oncogenetic Network Estimation with Disjunctive Bayesian Networks: Learning from Unstratified Samples while Preserving Mutual Exclusivity Relations

MotivationCancer is the process of accumulating genetic alterations that confer selective advantages to tumor cells. The order in which aberrations occur is not arbitrary, and inferring the order of events is challenging due to the lack of longitudinal samples from tumors. Moreover, a network model of oncogenesis should capture biological facts such as distinct progression trajectories of cancer subtypes and patterns of mutual exclusivity of alterations in the same pathways. In this paper, we present the Disjunctive Bayesian Network (DBN), a novel oncogenetic model with a phylogenetic interpretation. DBN is expressive enough to capture cancer subtypes trajectories and mutually exclusive relations between alterations from unstratified data. ResultsIn cases where the number of studied alterations is small (< 30), we provide an efficient dynamic programming implementation of an exact structure learning method that finds a best DBN in the super-exponential search space of networks. In rare cases that the number of alterations is large, we provided an efficient genetic algorithm in our software package, OncoBN. Through numerous synthetic and real data experiments, we show OncoBNs ability in inferring ground truth networks and recovering biologically meaningful progression networks. AvailabilityOncoBN is implemented in R and is available at https://github.com/phillipnicol/OncoBN.

bioinformatics↗

bioRxiv2019

Identification and comparison of genes differentially regulated by transcription factors and miRNAs

Transcription factors and microRNAs (miRNA) both play a critical role in gene regulation and in the development of many diseases such as cancer. Understanding how transcription factors and miRNAs influence gene expression is thus important to understand, but complicated due to the large and interconnected nature of the human genome. To help better understand what genes are being regulated by transcription factors and/or miRNAs we looked at it over 8000 patient samples from 32 different cancer types collected from The Cancer Genome Atlas (TCGA). We started by clustering the transcription factors and miRNAs using Thresher to reduce the number of features. Using both the mRNA and miRNA sequencing data we constructed linear models to calculate the coefficient of determination (R2) for each mRNA based on the Thresher cluster expression. We generated three types of linear models: transcription factor, miRNA and transcription factor plus miRNA. We then determined genes that were highly explained or poorly explained by each of the three models based on the genes R2 value. We performed downstream gene enrichment analysis using ToppGene on the sets of well and poorly explained genes. This identified differences in gene regulation between transcription factors and miRNAs and showed what groups of gene are differentially regulated.

systems biology↗