Search for: All records

Abstract Emerging studies underscore the promising capabilities of large language model-based chatbots in conducting basic bioinformatics data analyses. The recent feature of accepting image inputs by ChatGPT, also known as GPT-4V(ision), motivated us to explore its efficacy in deciphering bioinformatics scientific figures. Our evaluation with examples in cancer research, including sequencing data analysis, multimodal network-based drug repositioning, and tumor clonal evolution, revealed that ChatGPT can proficiently explain different plot types and apply biological knowledge to enrich interpretations. However, it struggled to provide accurate interpretations when color perception and quantitative analysis of visual elements were involved. Furthermore, while the chatbot can draft figure legends and summarize findings from the figures, stringent proofreading is imperative to ensure the accuracy and reliability of the content. 

There is currently no gene expression assay that can assess if premalignant lesions will develop into invasive breast cancer. This study sought to identify biomarkers for selecting patients with a high potential for developing invasive carcinoma in the breast with normal histology, benign lesions, or premalignant lesions. A set of 26-gene mRNA expression profiles were used to identify invasive ductal carcinomas from histologically normal tissue and benign lesions and to select those with a higher potential for future cancer development (ADHC) in the breast associated with atypical ductal hyperplasia (ADH). The expression-defined model achieved an overall accuracy of 94.05% (AUC = 0.96) in classifying invasive ductal carcinomas from histologically normal tissue and benign lesions (n = 185). This gene signature classified cancer development in ADH tissues with an overall accuracy of 100% (n = 8). The mRNA expression patterns of these 26 genes were validated using RT-PCR analyses of independent tissue samples (n = 77) and blood samples (n = 48). The protein expression of PBX2 and RAD52 assessed with immunohistochemistry were prognostic of breast cancer survival outcomes. This signature provided significant prognostic stratification in The Cancer Genome Atlas breast cancer patients (n = 1100), as well as basal-like and luminal A subtypes, and was associated with distinct immune infiltration and activities. The mRNA and protein expression of the 26 genes was associated with sensitivity or resistance to 18 NCCN-recommended drugs for treating breast cancer. Eleven genes had significant proliferative potential in CRISPR-Cas9/RNAi screening. Based on this gene expression signature, the VEGFR inhibitor ZM-306416 was discovered as a new drug for treating breast cancer. 

Breast cancer treatment can be improved with biomarkers for early detection and individualized therapy. A set of 86 microRNAs (miRNAs) were identified to separate breast cancer tumors from normal breast tissues (n = 52) with an overall accuracy of 90.4%. Six miRNAs had concordant expression in both tumors and breast cancer patient blood samples compared with the normal control samples. Twelve miRNAs showed concordant expression in tumors vs. normal breast tissues and patient survival (n = 1093), with seven as potential tumor suppressors and five as potential oncomiRs. From experimentally validated target genes of these 86 miRNAs, pan-sensitive and pan-resistant genes with concordant mRNA and protein expression associated with in-vitro drug response to 19 NCCN-recommended breast cancer drugs were selected. Combined with in-vitro proliferation assays using CRISPR-Cas9/RNAi and patient survival analysis, MEK inhibitors PD19830 and BRD-K12244279, pilocarpine, and tremorine were discovered as potential new drug options for treating breast cancer. Multi-omics biomarkers of response to the discovered drugs were identified using human breast cancer cell lines. This study presented an artificial intelligence pipeline of miRNA-based discovery of biomarkers, therapeutic targets, and repositioning drugs that can be applied to many cancer types. 

The majority of lung cancer patients are diagnosed with metastatic disease. This study identified a set of 73 microRNAs (miRNAs) that classified lung cancer tumors from normal lung tissues with an overall accuracy of 96.3% in the training patient cohort (n = 109) and 91.7% in unsupervised classification and 92.3% in supervised classification in the validation set (n = 375). Based on association with patient survival (n = 1016), 10 miRNAs were identified as potential tumor suppressors (hsa-miR-144, hsa-miR-195, hsa-miR-223, hsa-miR-30a, hsa-miR-30b, hsa-miR-30d, hsa-miR-335, hsa-miR-363, hsa-miR-451, and hsa-miR-99a), and 4 were identified as potential oncogenes (hsa-miR-21, hsa-miR-31, hsa-miR-411, and hsa-miR-494) in lung cancer. Experimentally confirmed target genes were identified for the 73 diagnostic miRNAs, from which proliferation genes were selected from CRISPR-Cas9/RNA interference (RNAi) screening assays. Pansensitive and panresistant genes to 21 NCCN-recommended drugs with concordant mRNA and protein expression were identified. DGKE and WDR47 were found with significant associations with responses to both systemic therapies and radiotherapy in lung cancer. Based on our identified miRNA-regulated molecular machinery, an inhibitor of PDK1/Akt BX-912, an anthracycline antibiotic daunorubicin, and a multi-targeted protein kinase inhibitor midostaurin were discovered as potential repositioning drugs for treating lung cancer. These findings have implications for improving lung cancer diagnosis, optimizing treatment selection, and discovering new drug options for better patient outcomes.