Next-Generation Machine Learning for Biological Networks.

Machine learning, a collection of data-analytical techniques aimed at building predictive models from multi-dimensional datasets, is becoming integral to modern biological research. By enabling one to generate models that learn from large datasets and make predictions on likely outcomes, machine learning can be used to study complex cellular systems such as biological networks. Here, we provide a primer on machine learning for life scientists, including an introduction to deep learning. We discuss opportunities and challenges at the intersection of machine learning and network biology, which could impact disease biology, drug discovery, microbiome research, and synthetic biology.

PRDM paralogs antagonistically balance Wnt/ß-catenin activity during craniofacial chondrocyte differentiation.

Cranial neural crest cell (NCC)-derived chondrocyte precursors undergo a dynamic differentiation and maturation process to establish a scaffold for subsequent bone formation, alterations in which contribute to congenital birth defects. Here, we demonstrate that transcription factor and histone methyltransferase proteins Prdm3 and Prdm16 control the differentiation switch of cranial NCCs to craniofacial cartilage. Loss of either paralog results in hypoplastic and disorganized chondrocytes due to impaired cellular orientation and polarity. We show that these proteins regulate cartilage differentiation by controlling the timing of Wnt/ß-catenin activity in strikingly different ways: Prdm3 represses whereas Prdm16 activates global gene expression, although both act by regulating Wnt enhanceosome activity and chromatin accessibility. Finally, we show that manipulating Wnt/ß-catenin signaling pharmacologically or generating prdm3-/-;prdm16-/- double mutants rescues craniofacial cartilage defects. Our findings reveal upstream regulatory roles for Prdm3 and Prdm16 in cranial NCCs to control Wnt/ß-catenin transcriptional activity during chondrocyte differentiation to ensure proper development of the craniofacial skeleton.

Crowdsourcing biomedical research: leveraging communities as innovation engines.

The generation of large-scale biomedical data is creating unprecedented opportunities for basic and translational science. Typically, the data producers perform initial analyses, but it is very likely that the most informative methods may reside with other groups. Crowdsourcing the analysis of complex and massive data has emerged as a framework to find robust methodologies. When the crowdsourcing is done in the form of collaborative scientific competitions, known as Challenges, the validation of the methods is inherently addressed. Challenges also encourage open innovation, create collaborative communities to solve diverse and important biomedical problems, and foster the creation and dissemination of well-curated data repositories.

High-Content Drug Discovery Targeting Molecular Bladder Cancer Subtypes.

Molecular subtypes of muscle-invasive bladder cancer (MIBC) display differential survival and drug sensitivities in clinical trials. To date, they have not been used as a paradigm for phenotypic drug discovery. This study aimed to discover novel subtype-stratified therapy approaches based on high-content screening (HCS) drug discovery. Transcriptome expression data of CCLE and BLA-40 cell lines were used for molecular subtype assignment in basal, luminal, and mesenchymal-like cell lines. Two independent HCSs, using focused compound libraries, were conducted to identify subtype-specific drug leads. We correlated lead drug sensitivity data with functional genomics, regulon analysis, and in-vitro drug response-based enrichment analysis. The basal MIBC subtype displayed sensitivity to HDAC and CHK inhibitors, while the luminal subtype was sensitive to MDM2 inhibitors. The mesenchymal-like cell lines were exclusively sensitive to the ITGAV inhibitor SB273005. The role of integrins within this mesenchymal-like MIBC subtype was confirmed via its regulon activity and gene essentiality based on CRISPR-Cas9 knock-out data. Patients with high ITGAV expression showed a significant decrease in the median overall survival. Phenotypic high-content drug screens based on bladder cancer cell lines provide rationales for novel stratified therapeutic approaches as a framework for further prospective validation in clinical trials.

Metastatic cells are preferentially vulnerable to lysosomal inhibition.

Molecular alterations that confer phenotypic advantages to tumors can also expose specific therapeutic vulnerabilities. To search for potential treatments that would selectively affect metastatic cells, we examined the sensitivity of lineage-related human bladder cancer cell lines with different lung colonization abilities to chloroquine (CQ) or bafilomycin A1, which are inhibitors of lysosome function and autophagy. Both CQ and bafilomycin A1 were more cytotoxic in vitro to highly metastatic cells compared with their less metastatic counterparts. Genetic inactivation of macroautophagy regulators and lysosomal proteins indicated that this was due to greater reliance on the lysosome but not upon macroautophagy. To identify the mechanism underlying these effects, we generated cells resistant to CQ in vitro. Surprisingly, selection for in vitro CQ resistance was sufficient to alter gene expression patterns such that unsupervised cluster analysis of whole-transcriptome data indicated that selection for CQ resistance alone created tumor cells that were more similar to the poorly metastatic parental cells from which the metastatic cells were derived; importantly, these tumor cells also had diminished metastatic ability in vivo. These effects were mediated in part by differential expression of the transcriptional regulator ID4 (inhibitor of DNA binding 4); depletion of ID4 both promoted in vitro CQ sensitivity and restored lung colonization and metastasis of CQ-resistant cells. These data demonstrate that selection for metastasis ability confers selective vulnerability to lysosomal inhibitors and identify ID4 as a potential biomarker for the use of lysosomal inhibitors to reduce metastasis in patients.

GSEA-InContext: identifying novel and common patterns in expression experiments.

Motivation: Gene Set Enrichment Analysis (GSEA) is routinely used to analyze and interpret coordinate pathway-level changes in transcriptomics experiments. For an experiment where less than seven samples per condition are compared, GSEA employs a competitive null hypothesis to test significance. A gene set enrichment score is tested against a null distribution of enrichment scores generated from permuted gene sets, where genes are randomly selected from the input experiment. Looking across a variety of biological conditions, however, genes are not randomly distributed with many showing consistent patterns of up- or down-regulation. As a result, common patterns of positively and negatively enriched gene sets are observed across experiments. Placing a single experiment into the context of a relevant set of background experiments allows us to identify both the common and experiment-specific patterns of gene set enrichment. Results: We compiled a compendium of 442 small molecule transcriptomic experiments and used GSEA to characterize common patterns of positively and negatively enriched gene sets. To identify experiment-specific gene set enrichment, we developed the GSEA-InContext method that accounts for gene expression patterns within a background set of experiments to identify statistically significantly enriched gene sets. We evaluated GSEA-InContext on experiments using small molecules with known targets to show that it successfully prioritizes gene sets that are specific to each experiment, thus providing valuable insights that complement standard GSEA analysis. Availability and implementation: GSEA-InContext implemented in Python, Supplementary results and the background expression compendium are available at: https://github.com/CostelloLab/GSEA-InContext.

Prediction of overall survival for patients with metastatic castration-resistant prostate cancer: development of a prognostic model through a crowdsourced challenge with open clinical trial data.

BACKGROUND: Improvements to prognostic models in metastatic castration-resistant prostate cancer have the potential to augment clinical trial design and guide treatment strategies. In partnership with Project Data Sphere, a not-for-profit initiative allowing data from cancer clinical trials to be shared broadly with researchers, we designed an open-data, crowdsourced, DREAM (Dialogue for Reverse Engineering Assessments and Methods) challenge to not only identify a better prognostic model for prediction of survival in patients with metastatic castration-resistant prostate cancer but also engage a community of international data scientists to study this disease. METHODS: Data from the comparator arms of four phase 3 clinical trials in first-line metastatic castration-resistant prostate cancer were obtained from Project Data Sphere, comprising 476 patients treated with docetaxel and prednisone from the ASCENT2 trial, 526 patients treated with docetaxel, prednisone, and placebo in the MAINSAIL trial, 598 patients treated with docetaxel, prednisone or prednisolone, and placebo in the VENICE trial, and 470 patients treated with docetaxel and placebo in the ENTHUSE 33 trial. Datasets consisting of more than 150 clinical variables were curated centrally, including demographics, laboratory values, medical history, lesion sites, and previous treatments. Data from ASCENT2, MAINSAIL, and VENICE were released publicly to be used as training data to predict the outcome of interest-namely, overall survival. Clinical data were also released for ENTHUSE 33, but data for outcome variables (overall survival and event status) were hidden from the challenge participants so that ENTHUSE 33 could be used for independent validation. Methods were evaluated using the integrated time-dependent area under the curve (iAUC). The reference model, based on eight clinical variables and a penalised Cox proportional-hazards model, was used to compare method performance. Further validation was done using data from a fifth trial-ENTHUSE M1-in which 266 patients with metastatic castration-resistant prostate cancer were treated with placebo alone. FINDINGS: 50 independent methods were developed to predict overall survival and were evaluated through the DREAM challenge. The top performer was based on an ensemble of penalised Cox regression models (ePCR), which uniquely identified predictive interaction effects with immune biomarkers and markers of hepatic and renal function. Overall, ePCR outperformed all other methods (iAUC 0·791; Bayes factor >5) and surpassed the reference model (iAUC 0·743; Bayes factor >20). Both the ePCR model and reference models stratified patients in the ENTHUSE 33 trial into high-risk and low-risk groups with significantly different overall survival (ePCR: hazard ratio 3·32, 95% CI 2·39-4·62, p<0·0001; reference model: 2·56, 1·85-3·53, p<0·0001). The new model was validated further on the ENTHUSE M1 cohort with similarly high performance (iAUC 0·768). Meta-analysis across all methods confirmed previously identified predictive clinical variables and revealed aspartate aminotransferase as an important, albeit previously under-reported, prognostic biomarker. INTERPRETATION: Novel prognostic factors were delineated, and the assessment of 50 methods developed by independent international teams establishes a benchmark for development of methods in the future. The results of this effort show that data-sharing, when combined with a crowdsourced challenge, is a robust and powerful framework to develop new prognostic models in advanced prostate cancer. FUNDING: Sanofi US Services, Project Data Sphere.

Differential splicing across immune system lineages.

Alternative splicing (AS) allows increased diversity and orthogonal regulation of the transcriptional products of mammalian genomes. To assess the distribution and variation of alternative splicing across cell lineages of the immune system, we comprehensively analyzed RNA sequencing and microarray data generated by the Immunological Genome Project Consortium. AS is pervasive: 60% of genes showed frequent AS isoforms in T or B lymphocytes, with 7,599 previously unreported isoforms. Distinct cell specificity was observed, with differential exon skipping in 5% of genes otherwise coexpressed in both B and T cells. The distribution of isoforms was mostly all or none, suggesting on/off switching as a frequent mode of AS regulation in lymphocytes. From the identification of differential exon use in the microarray data, clustering of exon inclusion/exclusion patterns across all Immunological Genome Project cell types showed that ∼70% of AS exons are distributed along a common pattern linked to lineage differentiation and cell cycling. Other AS events distinguished myeloid from lymphoid cells or affected only a small set of exons without clear lineage specificity (e.g., Ptprc). Computational analysis predicted specific associations between AS exons and splicing regulators, which were verified by detection of the hnRPLL/Ptprc connection.

Wisdom of crowds for robust gene network inference.

Reconstructing gene regulatory networks from high-throughput data is a long-standing challenge. Through the Dialogue on Reverse Engineering Assessment and Methods (DREAM) project, we performed a comprehensive blind assessment of over 30 network inference methods on Escherichia coli, Staphylococcus aureus, Saccharomyces cerevisiae and in silico microarray data. We characterize the performance, data requirements and inherent biases of different inference approaches, and we provide guidelines for algorithm application and development. We observed that no single inference method performs optimally across all data sets. In contrast, integration of predictions from multiple inference methods shows robust and high performance across diverse data sets. We thereby constructed high-confidence networks for E. coli and S. aureus, each comprising ~1,700 transcriptional interactions at a precision of ~50%. We experimentally tested 53 previously unobserved regulatory interactions in E. coli, of which 23 (43%) were supported. Our results establish community-based methods as a powerful and robust tool for the inference of transcriptional gene regulatory networks.

ProstaMine: a bioinformatics tool for identifying subtype-specific co-alterations associated with aggressiveness in prostate cancer.

Background: Prostate cancer is a leading cause of cancer-related deaths among men, marked by heterogeneous clinical and molecular characteristics. The complexity of the molecular landscape necessitates tools for identifying multi-gene co-alteration patterns that are associated with aggressive disease. The identification of such gene sets will allow for deeper characterization of the processes underlying prostate cancer progression and potentially lead to novel strategies for treatment. Methods: We developed ProstaMine to systematically identify co-alterations associated with aggressiveness in prostate cancer molecular subtypes defined by high-fidelity alterations in primary prostate cancer. ProstaMine integrates genomic, transcriptomic, and clinical data from five primary and one metastatic prostate cancer cohorts to prioritize co-alterations enriched in metastatic disease and associated with disease progression. Results: Integrated analysis of primary tumors defined a set of 17 prostate cancer alterations associated with aggressive characteristics. We applied ProstaMine to NKX3-1-loss and RB1-loss tumors and identified subtype-specific co-alterations associated with metastasis and biochemical relapse in these molecular subtypes. In NKX3-1-loss prostate cancer, ProstaMine identified novel subtype-specific co-alterations known to regulate prostate cancer signaling pathways including MAPK, NF-kB, p53, PI3K, and Sonic hedgehog. In RB1-loss prostate cancer, ProstaMine identified novel subtype-specific co-alterations involved in p53, STAT6, and MHC class I antigen presentation. Co-alterations impacting autophagy were noted in both molecular subtypes. Conclusion: ProstaMine is a method to systematically identify novel subtype-specific co-alterations associated with aggressive characteristics in prostate cancer. The results from ProstaMine provide insights into potential subtype-specific mechanisms of prostate cancer progression which can be formed into testable experimental hypotheses. ProstaMine is publicly available at: https://bioinformatics.cuanschutz.edu/prostamine.

E-cadherin expression and gene expression profiles in corticotroph pituitary neuroendocrine tumor subtypes.

Corticotroph adenomas/pituitary neuroendocrine tumors (PitNETs) are associated with significant morbidity and mortality. Predictors of tumor behavior have not shown high prognostic accuracy. For somatotroph adenomas/PitNETs, E-cadherin expression correlates strongly with prognosis. E-cadherin expression has not been investigated in other PitNETs. A retrospective chart review of adults with corticotroph adenomas/PitNETs was conducted to assess correlation between E-cadherin expression and tumor characteristics. In addition, gene expression microarray was performed in subset of tumors (n = 16). Seventy-seven patients were identified; 71% were female, with median age of cohort 45.2 years. Seventy-five percent had macroadenomas, of which 22% were hormonally active. Ninety-five percent of microadenomas were hormonally active. Adrenocorticotropic hormone granulation pattern by IHC identified 63% as densely granulated (DG) and 34% as sparsely granulated (SG). All microadenomas were DG (p < .001); 50% of macroadenomas were DG associated with increased tumor invasion compared to SG. E-cadherin IHC was positive in 80%, diminished in 17%, and absent in 20% and did not correlate with corticotroph PitNETs subtype, size, or prognosis. In contrast to the distinct transcriptomes of corticotroph PitNETs and normal pituitaries, a comparison of clinically active and silent corticotroph PitNETs demonstrated similar molecular signatures indicating their common origin, but with unique differences related to their secretory status.

NPEPPS Is a Druggable Driver of Platinum Resistance.

There is an unmet need to improve the efficacy of platinum-based cancer chemotherapy, which is used in primary and metastatic settings in many cancer types. In bladder cancer, platinum-based chemotherapy leads to better outcomes in a subset of patients when used in the neoadjuvant setting or in combination with immunotherapy for advanced disease. Despite such promising results, extending the benefits of platinum drugs to a greater number of patients is highly desirable. Using the multiomic assessment of cisplatin-responsive and -resistant human bladder cancer cell lines and whole-genome CRISPR screens, we identified puromycin-sensitive aminopeptidase (NPEPPS) as a driver of cisplatin resistance. NPEPPS depletion sensitized resistant bladder cancer cells to cisplatin in vitro and in vivo. Conversely, overexpression of NPEPPS in sensitive cells increased cisplatin resistance. NPEPPS affected treatment response by regulating intracellular cisplatin concentrations. Patient-derived organoids (PDO) generated from bladder cancer samples before and after cisplatin-based treatment, and from patients who did not receive cisplatin, were evaluated for sensitivity to cisplatin, which was concordant with clinical response. In the PDOs, depletion or pharmacologic inhibition of NPEPPS increased cisplatin sensitivity, while NPEPPS overexpression conferred resistance. Our data present NPEPPS as a druggable driver of cisplatin resistance by regulating intracellular cisplatin concentrations. SIGNIFICANCE: Targeting NPEPPS, which induces cisplatin resistance by controlling intracellular drug concentrations, is a potential strategy to improve patient responses to platinum-based therapies and lower treatment-associated toxicities.

curatedPCaData: Integration of clinical, genomic, and signature features in a curated and harmonized prostate cancer data resource.

Genomic and transcriptomic data have been generated across a wide range of prostate cancer (PCa) study cohorts. These data can be used to better characterize the molecular features associated with clinical outcomes and to test hypotheses across multiple, independent patient cohorts. In addition, derived features, such as estimates of cell composition, risk scores, and androgen receptor (AR) scores, can be used to develop novel hypotheses leveraging existing multi-omic datasets. The full potential of such data is yet to be realized as independent datasets exist in different repositories, have been processed using different pipelines, and derived and clinical features are often not provided or unstandardized. Here, we present the curatedPCaData R package, a harmonized data resource representing >2900 primary tumor, >200 normal tissue, and >500 metastatic PCa samples across 19 datasets processed using standardized pipelines with updated gene annotations. We show that meta-analysis across harmonized studies has great potential for robust and clinically meaningful insights. curatedPCaData is an open and accessible community resource with code made available for reproducibility.

SIX1 and EWS/FLI1 co-regulate an anti-metastatic gene network in Ewing Sarcoma.

Ewing sarcoma (ES), which is characterized by the presence of oncogenic fusion proteins such as EWS/FLI1, is an aggressive pediatric malignancy with a high rate of early dissemination and poor outcome after distant spread. Here we demonstrate that the SIX1 homeoprotein, which enhances metastasis in most tumor types, suppresses ES metastasis by co-regulating EWS/FLI1 target genes. Like EWS/FLI1, SIX1 promotes cell growth/transformation, yet dramatically inhibits migration and invasion, as well as metastasis in vivo. We show that EWS/FLI1 promotes SIX1 protein expression, and that the two proteins share genome-wide binding profiles and transcriptional regulatory targets, including many metastasis-associated genes such as integrins, which they co-regulate. We further show that SIX1 downregulation of integrins is critical to its ability to inhibit invasion, a key characteristic of metastatic cells. These data demonstrate an unexpected anti-metastatic function for SIX1, through coordinate gene regulation with the key oncoprotein in ES, EWS/FLI1.

A harmonized resource of integrated prostate cancer clinical, -omic, and signature features.

Genomic and transcriptomic data have been generated across a wide range of prostate cancer (PCa) study cohorts. These data can be used to better characterize the molecular features associated with clinical outcomes and to test hypotheses across multiple, independent patient cohorts. In addition, derived features, such as estimates of cell composition, risk scores, and androgen receptor (AR) scores, can be used to develop novel hypotheses leveraging existing multi-omic datasets. The full potential of such data is yet to be realized as independent datasets exist in different repositories, have been processed using different pipelines, and derived and clinical features are often not provided or  not standardized. Here, we present the curatedPCaData R package, a harmonized data resource representing >2900 primary tumor, >200 normal tissue, and >500 metastatic PCa samples across 19 datasets processed using standardized pipelines with updated gene annotations. We show that meta-analysis across harmonized studies has great potential for robust and clinically meaningful insights. curatedPCaData is an open and accessible community resource with code made available for reproducibility.

Microparticle-Delivered Cxcl9 Prolongs Braf Inhibitor Efficacy in Melanoma.

Patients with BRAF-mutant melanoma show substantial responses to combined BRAF and MEK inhibition, but most relapse within 2 years. A major reservoir for drug resistance is minimal residual disease (MRD), comprised of drug-tolerant tumor cells laying in a dormant state. Towards exploiting potential therapeutic vulnerabilities of MRD, we established a genetically engineered mouse model of BrafV600E-driven melanoma MRD wherein genetic BrafV600E extinction leads to strong but incomplete tumor regression. Transcriptional time-course analysis after BrafV600E extinction revealed that after an initial surge of immune activation, tumors later became immunologically "cold" after MRD establishment. Computational analysis identified candidate T-cell recruiting chemokines as strongly upregulated initially and steeply decreasing as the immune response faded. Therefore, we hypothesized that sustaining chemokine signaling could impair MRD maintenance through increased recruitment of effector T cells. We found that intratumoral administration of recombinant Cxcl9 (rCxcl9), either naked or loaded in microparticles, significantly impaired MRD relapse in BRAF-inhibited tumors, including several complete pathologic responses after microparticle-delivered rCxcl9 combined with BRAF and MEK inhibition. Our experiments constitute proof of concept that chemokine-based microparticle delivery systems are a potential strategy to forestall tumor relapse and thus improve the clinical success of first-line treatment methods.

Down syndrome is associated with altered frequency and functioning of tracheal multiciliated cells, and response to influenza virus infection.

Individuals with Down syndrome (DS) clinically manifest severe respiratory illnesses; however, there is a paucity of data on how DS influences homeostatic physiology of lung airway, and its reactive responses to pulmonary pathogens. We generated well-differentiated ciliated airway epithelia using tracheas from wild-type and Dp(16)1/Yey mice in vitro, and discovered that Dp(16)1/Yey epithelia have significantly lower abundance of ciliated cells, an altered ciliary beating profile, and reduced mucociliary transport. Interestingly, both sets of differentiated epithelia released similar quantities of viral particles after infection with influenza A virus (IAV). However, RNA-sequencing and proteomic analyses revealed an immune hyperreactive phenotype particularly for monocyte-recruiting chemokines in Dp(16)1/Yey epithelia. Importantly, when we challenged mice in vivo with IAV, we observed immune hyper-responsiveness in Dp(16)1/Yey mice, evidenced by higher quantities of lung airway infiltrated monocytes, and elevated levels of pro-inflammatory cytokines in bronchoalveolar lavage fluid. Our findings illuminate mechanisms underlying DS-mediated pathophysiological changes in airway epithelium.

Non-muscle-invasive bladder cancer molecular subtypes predict differential response to intravesical Bacillus Calmette-Guérin.

The recommended treatment for patients with high-risk non-muscle-invasive bladder cancer (HR-NMIBC) is tumor resection followed by adjuvant Bacillus Calmette-Guérin (BCG) bladder instillations. However, only 50% of patients benefit from this therapy. If progression to advanced disease occurs, then patients must undergo a radical cystectomy with risks of substantial morbidity and poor clinical outcome. Identifying tumors unlikely to respond to BCG can translate into alternative treatments, such as early radical cystectomy, targeted therapies, or immunotherapies. Here, we conducted molecular profiling of 132 patients with BCG-naive HR-NMIBC and 44 patients with recurrences after BCG (34 matched), which uncovered three distinct BCG response subtypes (BRS1, 2 and BRS3). Patients with BRS3 tumors had a reduced recurrence-free and progression-free survival compared with BRS1/2. BRS3 tumors expressed high epithelial-to-mesenchymal transition and basal markers and had an immunosuppressive profile, which was confirmed with spatial proteomics. Tumors that recurred after BCG were enriched for BRS3. BRS stratification was validated in a second cohort of 151 BCG-naive patients with HR-NMIBC, and the molecular subtypes outperformed guideline-recommended risk stratification based on clinicopathological variables. For clinical application, we confirmed that a commercially approved assay was able to predict BRS3 tumors with an area under the curve of 0.87. These BCG response subtypes will allow for improved identification of patients with HR-NMIBC at the highest risk of progression and have the potential to be used to select more appropriate treatments for patients unlikely to respond to BCG.

Microbiome Preterm Birth DREAM Challenge: Crowdsourcing Machine Learning Approaches to Advance Preterm Birth Research.

Globally, every year about 11% of infants are born preterm, defined as a birth prior to 37 weeks of gestation, with significant and lingering health consequences. Multiple studies have related the vaginal microbiome to preterm birth. We present a crowdsourcing approach to predict: (a) preterm or (b) early preterm birth from 9 publicly available vaginal microbiome studies representing 3,578 samples from 1,268 pregnant individuals, aggregated from raw sequences via an open-source tool, MaLiAmPi. We validated the crowdsourced models on novel datasets representing 331 samples from 148 pregnant individuals. From 318 DREAM challenge participants we received 148 and 121 submissions for our two separate prediction sub-challenges with top-ranking submissions achieving bootstrapped AUROC scores of 0.69 and 0.87, respectively. Alpha diversity, VALENCIA community state types, and composition (via phylotype relative abundance) were important features in the top performing models, most of which were tree based methods. This work serves as the foundation for subsequent efforts to translate predictive tests into clinical practice, and to better understand and prevent preterm birth.

SOPHIE: Generative Neural Networks Separate Common and Specific Transcriptional Responses.

Genome-wide transcriptome profiling identifies genes that are prone to differential expression (DE) across contexts, as well as genes with changes specific to the experimental manipulation. Distinguishing genes that are specifically changed in a context of interest from common differentially expressed genes (DEGs) allows more efficient prediction of which genes are specific to a given biological process under scrutiny. Currently, common DEGs or pathways can only be identified through the laborious manual curation of experiments, an inordinately time-consuming endeavor. Here we pioneer an approach, Specific cOntext Pattern Highlighting In Expression data (SOPHIE), for distinguishing between common and specific transcriptional patterns using a generative neural network to create a background set of experiments from which a null distribution of gene and pathway changes can be generated. We apply SOPHIE to diverse datasets including those from human, human cancer, and bacterial pathogen Pseudomonas aeruginosa. SOPHIE identifies common DEGs in concordance with previously described, manually and systematically determined common DEGs. Further molecular validation indicates that SOPHIE detects highly specific but low-magnitude biologically relevant transcriptional changes. SOPHIE's measure of specificity can complement log2 fold change values generated from traditional DE analyses. For example, by filtering the set of DEGs, one can identify genes that are specifically relevant to the experimental condition of interest. Consequently, these results can inform future research directions. All scripts used in these analyses are available at https://github.com/greenelab/generic-expression-patterns. Users can access https://github.com/greenelab/sophie to run SOPHIE on their own data.