Pathformer: a biological pathway informed transformer for disease diagnosis and prognosis using multi-omics data.

MOTIVATION: Multi-omics data provide a comprehensive view of gene regulation at multiple levels, which is helpful in achieving accurate diagnosis of complex diseases like cancer. However, conventional integration methods rarely utilize prior biological knowledge and lack interpretability. RESULTS: To integrate various multi-omics data of tissue and liquid biopsies for disease diagnosis and prognosis, we developed a biological pathway informed Transformer, Pathformer. It embeds multi-omics input with a compacted multi-modal vector and a pathway-based sparse neural network. Pathformer also leverages criss-cross attention mechanism to capture the crosstalk between different pathways and modalities. We first benchmarked Pathformer with 18 comparable methods on multiple cancer datasets, where Pathformer outperformed all the other methods, with an average improvement of 6.3%-14.7% in F1 score for cancer survival prediction, 5.1%-12% for cancer stage prediction, and 8.1%-13.6% for cancer drug response prediction. Subsequently, for cancer prognosis prediction based on tissue multi-omics data, we used a case study to demonstrate the biological interpretability of Pathformer by identifying key pathways and their biological crosstalk. Then, for cancer early diagnosis based on liquid biopsy data, we used plasma and platelet datasets to demonstrate Pathformer's potential of clinical applications in cancer screening. Moreover, we revealed deregulation of interesting pathways (e.g. scavenger receptor pathway) and their crosstalk in cancer patients' blood, providing potential candidate targets for cancer microenvironment study. AVAILABILITY AND IMPLEMENTATION: Pathformer is implemented and freely available at https://github.com/lulab/Pathformer.

Evaluating the Utilities of Foundation Models in Single-cell Data Analysis.

Foundation Models (FMs) have made significant strides in both industrial and scientific domains. In this paper, we evaluate the performance of FMs in single-cell sequencing data analysis through comprehensive experiments across eight downstream tasks pertinent to single-cell data. By comparing ten different single-cell FMs with task-specific methods, we found that single-cell FMs may not consistently excel in all tasks than task-specific methods. However, the emergent abilities and the successful applications of cross-species/cross-modality transfer learning of FMs are promising. In addition, we present a systematic evaluation of the effects of hyper-parameters, initial settings, and stability for training single-cell FMs based on a proposed scEval framework, and provide guidelines for pre-training and fine-tuning. Our work summarizes the current state of single-cell FMs and points to their constraints and avenues for future development.

Constituent analysis, laxative activity, and toxicological evaluation of methanol extract of noni fruit (Morinda citrifolia L., Rubiaceae).

Noni fruits have gained considerable popularity as dietary supplements. However, the major constituents, the laxative activity, and the toxicity of Noni fruit remains still unknown. The purpose of the present study was, therefore, to analyze the constituents of methanol extract of Noni fruit by UPLC-MS, and further evaluate laxative activity and safety aspects of this Noni fruit-derived products in mice. UPLC-MS analysis identified eleven major constituents from this Noni fruit extract. Administration of this extract obviously shortened the time of first fecal excrement, significantly increased the total number and the weight of stools, and remarkably restored the intestinal transit to normal level in the constipated mice, with low toxicity to liver and kidney, and meanwhile, the abundance, composition, and function of gut microbiota remained homeostasis. These results revealed the laxative activity of the methanol extract of Noni fruit with low toxicity and no influence on gut microbiota.

Differential expression spectrum and targeted gene prediction of tRNA-derived small RNAs in idiopathic pulmonary arterial hypertension.

Background: Idiopathic pulmonary arterial hypertension (PAH) is a potentially fatal pulmonary vascular disease with an extremely poor natural course. The limitations of current treatment and the unclear etiology and pathogenesis of idiopathic PAH require new targets and avenues of exploration involved in the pathogenesis of PAH. tRNA-derived small RNAs (tsRNAs), a new type of small non-coding RNAs, have a significant part in the progress of diverse diseases. However, the potential functions behind tsRNAs in idiopathic PAH remain unknown. Methods: Small RNA microarray was implemented on three pairs of plasma of idiopathic PAH patients and healthy controls to investigate and compare tsRNAs expression profiles. Validation samples were used for real-time polymerase chain reaction (Real-time PCR) to verify several dysregulated tsRNAs. Bioinformatic analysis was adopted to determine potential target genes and mechanisms of the validated tsRNAs in PAH. Results: Microarray detected 816 statistically significantly dysregulated tsRNAs, of which 243 tsRNAs were upregulated and 573 were downregulated in PAH. Eight validated tsRNAs in the results of Real-time PCR were concordant with the small RNA microarray: four upregulated (tRF3a-AspGTC-9, 5'tiRNA-31-GluCTC-16, i-tRF-31:54-Val-CAC-1 and tRF3b-TyrGTA-4) and four downregulated (5'tiRNA-33-LysTTT-4, i-tRF-8:32-Val-AAC-2, i-tRF-2:30-His-GTG-1, and i-tRF-15:31-Lys-CTT-1). The Gene Ontology analysis has shown that the verified tsRNAs are related to cellular macromolecule metabolic process, regulation of cellular process, and regulation of cellular metabolic process. It is disclosed that potential target genes of verified tsRNAs are widely involved in PAH pathways by Kyoto Encyclopedia of Genes and Genomes. Conclusion: This study investigated tsRNA profiles in idiopathic PAH and found that the dysregulated tsRNAs may become a novel type of biomarkers and possible targets for PAH.