Semi-supervised methods for single-cell RNA-seq integration promise improved batch correction and preservation of biological signal by leveraging cell-type labels. However, reported benefits and robustness of them towards imperfect cell type labels often come from overly idealized settings. Here we present, to our knowledge, the first systematic benchmark comparing leading semi-supervised methods with widely used unsupervised approaches across six diverse datasets under realistic conditions. Beyond randomly missing or erroneous labels, we examine four additional scenarios (boundary-mixed labels, batch-specific annotations, auto-generated labels and varied-granularity labels) and evaluate performance using nine established metrics. We find that although semi-supervised methods can provide benefits under perfect annotations, their robustness often degrades substantially under realistic imperfections. Only scANVI and ssSTACAS maintain stable but modest improvements over their unsupervised counterparts, and none consistently outperform the strongest unsupervised approach. These results indicate that current semi-supervised strategies offer limited practical advantage when label quality is modest uncertain.

*Equalized odds*, an important notion of algorithmic fairness, aims to ensure that sensitive variables, such as race and gender, do not unfairly influence the algorithm's prediction when conditioning on the true outcome. Despite rapid advancements, current research primarily focuses on equalized odds violations caused by a single sensitive attribute, leaving the challenge of simultaneously accounting for multiple attributes under-addressed. We bridge this gap by introducing an in-processing fairness-aware learning approach, FairICP, which integrates adversarial learning with a novel inverse conditional permutation scheme. FairICP offers a flexible and efficient scheme to promote equalized odds under fairness conditions described by complex and multi-dimensional sensitive attributes. The efficacy and adaptability of our method are demonstrated through both simulation studies and empirical analyses of real-world datasets.

Systems vaccinology studies have been used to build computational models that predict individual vaccine responses and identify the factors contributing to differences in outcome. Comparing such models is challenging due to variability in study designs. To address this, we established a community resource to compare models predicting B. pertussis booster responses and generate experimental data for the explicit purpose of model evaluation. We here describe our second computational prediction challenge using this resource, where we benchmarked 49 algorithms from 53 scientists. We found that the most successful models stood out in their handling of nonlinearities, reducing large feature sets to representative subsets, and advanced data preprocessing. In contrast, we found that models adopted from literature that were developed to predict vaccine antibody responses in other settings performed poorly, reinforcing the need for purpose-built models. Overall, this demonstrates the value of purpose-generated datasets for rigorous and open model evaluations to identify features that improve the reliability and applicability of computational models in vaccine response prediction.

The Random Forests classifier, a widely utilized off-the-shelf classification tool, assumes training and test samples come from the same distribution as other standard classifiers. However, in safety-critical scenarios like medical diagnosis and network attack detection, discrepancies between the training and test sets, including the potential presence of novel outlier samples not appearing during training, can pose significant challenges. To address this problem, we introduce the Conformalized Semi-Supervised Random Forest (CSForest), which couples the conformalization technique Jackknife+aB with semi-supervised tree ensembles to construct a set-valued prediction $C(x)$. Instead of optimizing over the training distribution, CSForest employs unlabeled test samples to enhance accuracy and flag unseen outliers by generating an empty set. Theoretically, we establish CSForest to cover true labels for previously observed inlier classes under arbitrarily label-shift in the test data. We compare CSForest with state-of-the-art methods using synthetic examples and various real-world datasets, under different types of distribution changes in the test domain. Our results highlight CSForest’s effective prediction of inliers and its ability to detect outlier samples unique to the test data. In addition, CSForest shows persistently good performance as the sizes of the training and test sets vary. Codes of CSForest are available at https://github.com/yujinhan98/CSForest.

MOTIVATION: Predictive biological signatures provide utility as biomarkers for disease diagnosis and prognosis, as well as prediction of responses to vaccination or therapy. These signatures are identified from high-throughput profiling assays through a combination of dimensionality reduction and machine learning techniques. The genes, proteins, metabolites, and other biological analytes that compose signatures also generate hypotheses on the underlying mechanisms driving biological responses, thus improving biological understanding. Dimensionality reduction is a critical step in signature discovery to address the large number of analytes in omics datasets, especially for multi-omics profiling studies with tens of thousands of measurements. Latent factor models, which can account for the structural heterogeneity across diverse assays, effectively integrate multi-omics data and reduce dimensionality to a small number of factors that capture correlations and associations among measurements. These factors provide biologically interpretable features for predictive modeling. However, multi-omics integration and predictive modeling are generally performed independently in sequential steps, leading to suboptimal factor construction. Combining these steps can yield better multi-omics signatures that are more predictive while still being biologically meaningful. RESULTS: We developed a supervised variational Bayesian factor model that extracts multi-omics signatures from high-throughput profiling datasets that can span multiple data types. Signature-based multiPle-omics intEgration via lAtent factoRs (SPEAR) adaptively determines factor rank, emphasis on factor structure, data relevance and feature sparsity. The method improves the reconstruction of underlying factors in synthetic examples and prediction accuracy of coronavirus disease 2019 severity and breast cancer tumor subtypes. AVAILABILITY AND IMPLEMENTATION: SPEAR is a publicly available R-package hosted at https://bitbucket.org/kleinstein/SPEAR.

Mass cytometry (CyTOF) has greatly expanded the capability of cytometry. It is now easy to generate multiple CyTOF samples in a single study, with each sample containing single-cell measurement on 50 markers for more than hundreds of thousands of cells. Current methods do not adequately address the issues concerning combining multiple samples for subpopulation discovery, and these issues can be quickly and dramatically amplified with increasing number of samples. To overcome this limitation, we developed Partition-Assisted Clustering and Multiple Alignments of Networks (PAC-MAN) for the fast automatic identification of cell populations in CyTOF data closely matching that of expert manual-discovery, and for alignments between subpopulations across samples to define dataset-level cellular states. PAC-MAN is computationally efficient, allowing the management of very large CyTOF datasets, which are increasingly common in clinical studies and cancer studies that monitor various tissue samples for each subject.