Accurate prediction of gene expression levels across different tissues is of great significance in understanding the functional roles of genes in various biological processes and assisting in transcriptome-wide association studies (TWAS). Traditional methods rely on the construction of a prediction model for each gene in a specific tissue, which is time-consuming and inefficient when dealing with numerous genes or tissues. In addition, current approaches do not consider missing single nucleotide polymorphisms (SNPs) in the training population. These SNPs significantly affect gene expression levels and in turn limit the predictive capability of these approaches in new samples. Recent research indicates that by identifying a specific group of core genes (known as landmark genes) that accurately reflect the cellular states of samples across different experimental conditions, it is possible to predict the expression levels of other genes in the genome. In light of this, we propose AGEP_TWAS (Adaptive Gene Expression Predictor for TWAS), a gene expression prediction method that utilizes a dense connection network, adaptive activation functions, and parameter pruning strategies within a nonlinear feature extraction computational framework. AGEP_TWAS leverages landmark genes within a tissue to predict the expression levels of other genes that are challenging to predict using traditional methods. Results on the human GEO expression dataset demonstrate that AGEP_TWAS achieved a mean squared error (MSE) of 0.1821 and a Pearson correlation coefficient (PCC) of 0.9004, outperforming existing state-of-the-art prediction models. Additionally, when applied to the CattleGTEx dataset to infer gene expression levels across different tissues in cattle, AGEP_TWAS exhibited superior predictive performance compared to existing methods. A TWAS on milk production traits in cattle highlights the practical utility of AGEP_TWAS, with six identified significant genes already reported in scientific literature to be associated with milk production traits.

MOTIVATION: Effective identification of cell types is of critical importance in single-cell RNA-sequencing (scRNA-seq) data analysis. To date, many supervised machine learning-based predictors have been implemented to identify cell types from scRNA-seq datasets. Despite the technical advances of these state-of-the-art tools, most existing predictors were single classifiers, of which the performances can still be significantly improved. It is therefore highly desirable to employ the ensemble learning strategy to develop more accurate computational models for robust and comprehensive identification of cell types on scRNA-seq datasets. RESULTS: We propose a two-layer stacking model, termed CTISL (Cell Type Identification by Stacking ensemble Learning), which integrates multiple classifiers to identify cell types. In the first layer, given a reference scRNA-seq dataset with known cell types, CTISL dynamically combines multiple cell-type-specific classifiers (i.e. support-vector machine and logistic regression) as the base learners to deliver the outcomes for the input of a meta-classifier in the second layer. We conducted a total of 24 benchmarking experiments on 17 human and mouse scRNA-seq datasets to evaluate and compare the prediction performance of CTISL and other state-of-the-art predictors. The experiment results demonstrate that CTISL achieves superior or competitive performance compared to these state-of-the-art approaches. We anticipate that CTISL can serve as a useful and reliable tool for cost-effective identification of cell types from scRNA-seq datasets. AVAILABILITY AND IMPLEMENTATION: The webserver and source code are freely available at http://bigdata.biocie.cn/CTISLweb/home and https://zenodo.org/records/10568906, respectively.