A comprehensive evaluation of full-spectrum cell-free RNAs highlights cell-free RNA fragments for early-stage hepatocellular carcinoma detection

Ning, Chun; Cai, Peng; Liu, Xiaofan; Li, Guangtao; Bao, Pengfei; Yan, Ling; Meng, Na; Tang, Kaichen; Luo, Yi; Guo, Hua; Wang, Yunjiu; Wang, Zhuoran; Chen, Lu; Lu, Zhi John; Yin, Jianhua

doi:10.1016/j.ebiom.2023.104645

Cited by 10 publications

(3 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Promising applications for diagnosis and prognosis of cfRNA has been demonstrated in a variety of cancers, including liver cancer, myeloma, colorectal cancer, breast cancer, lung cancer, etc. [78][79][80][81] A number of differential and prognostic IR events with tissue specificity were found in our study, some of which may have the potential to predict tumor tissue origin. Through subsequent rigorous screening and validation in plasma from tumor and normal samples, IR biomarkers may be applicable in liquid biopsy.…”

Section: Discussionmentioning

confidence: 58%

Pan-Cancer Profiling of Intron Retention and Its Clinical Significance in Diagnosis and Prognosis

Huang,

Zeng,

et al. 2023

Cancers

View full text Add to dashboard Cite

Alternative splicing can produce transcripts that affect cancer development and thus shows potential for cancer diagnosis and treatment. However, intron retention (IR), a type of alternative splicing, has been studied less in cancer biology research. Here, we generated a pan-cancer IR landscape for more than 10,000 samples across 33 cancer types from The Cancer Genome Atlas (TCGA). We characterized differentially retained introns between tumor and normal samples and identified retained introns associated with survival. We discovered 988 differentially retained introns in 14 cancers, some of which demonstrated diagnostic potential in multiple cancer types. We also inferred a large number of prognosis-related introns in 33 cancer types, and the associated genes included well-known cancer hallmarks such as angiogenesis, metastasis, and DNA mutations. Notably, we discovered a novel intron retention inside the 5′UTR of STN1 that is associated with the survival of lung cancer patients. The retained intron reduces translation efficiency by producing upstream open reading frames (uORFs) and thereby inhibits colony formation and cell migration of lung cancer cells. Besides, the IR-based prognostic model achieved good stratification in certain cancers, as illustrated in acute myeloid leukemia. Taken together, we performed a comprehensive IR survey at a pan-cancer level, and the results implied that IR has the potential to be diagnostic and prognostic cancer biomarkers, as well as new drug targets.

show abstract

Section: Discussionmentioning

confidence: 58%

Pan-Cancer Profiling of Intron Retention and Its Clinical Significance in Diagnosis and Prognosis

Huang,

Zeng,

et al. 2023

Cancers

View full text Add to dashboard Cite

show abstract

“…Comparing to a single type of data, multi-omics data provide a more comprehensive view of gene regulation ( Hasin et al 2017 ). Therefore, integrating multi-omics data from tissue and liquid biopsies would be helpful in addressing challenges in disease diagnosis ( Ning et al 2023 ), treatment ( Chiu et al 2019 , Sharifi-Noghabi et al 2019 ), and prognosis ( Hao et al 2018 ), such as deregulated network between different types of molecules and data noise caused by patients’ heterogeneity ( Tarazona et al 2021 ). To integrate multi-omics data of cancer, several supervised methods have been developed, such as mixOmics ( Rohart et al 2017 ), liNN ( Kuru et al 2022 ), eiNN ( Preuer et al 2018 ), liCNN ( Islam et al 2020 ), eiCNN ( Fu et al 2020 ), MOGONet ( Wang et al 2021 ), and MOGAT ( Xing et al 2021 ).…”

Section: Introductionmentioning

confidence: 99%

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

Liu,

Tao,

Cai

et al. 2024

Bioinformatics

Self Cite

View full text Add to dashboard Cite

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 Pathformer is implemented and freely available at https://github.com/lulab/Pathformer. Supplementary information Supplementary data are available at Bioinformatics online.

show abstract

“…Nonetheless, the field faces challenges in sensitivity and specificity due to variability in cfRNA levels and the difficulty in distinguishing cancerous cfRNA profiles from normal ones. To tackle this, studies have integrated machine learning methods, typically performing Differentially Expression (DE) analysis to discover Differentially Expressed Genes (DEGs) and identify potential biomarkers using methods such as random forest [29]. However, these methods are still reliant on known genes.…”

mentioning

confidence: 99%

GeneLLM: A Large cfRNA Language Model for Cancer Screening from Raw Reads

Deng,

Sha,

Jin

et al. 2024

Preprint

View full text Add to dashboard Cite

Plasma cell-free RNA (cfRNA) has recently emerged as a promising biomarker for non-invasive early cancer detection and treatment monitoring [8, 28, 31, 32, 38, 40]. Here, we introduce GeneLLM, a novel large language model de-signed to interpret cfRNA sequences directly, bypassing the need for genome annotations. GeneLLM significantly advances the detection accuracy of various cancer types. Our study demonstrates that this method achieves higher accuracy than traditional biomarkers and effectively handles large datasets from different centres, even with low sequencing depth. By avoiding the use of bioinformatics tools to count known genes, GeneLLM also discovered cfRNAs from previously unknown genes, referred to as ‘dark matters’ in the genome, as cancer detec-tion ‘pseudo-biomarkers’. Our results showcase the potential of GeneLLM to revolutionise cancer detection, making it more accessible and cost-effective. By offering a method that does not depend on bioinformatics tools to count known genes, GeneLLM opens new avenues for biomarker discovery and enhances our understanding of intercellular communication through novel RNA molecules.

show abstract

A comprehensive evaluation of full-spectrum cell-free RNAs highlights cell-free RNA fragments for early-stage hepatocellular carcinoma detection

Cited by 10 publications

References 51 publications

Pan-Cancer Profiling of Intron Retention and Its Clinical Significance in Diagnosis and Prognosis

Pan-Cancer Profiling of Intron Retention and Its Clinical Significance in Diagnosis and Prognosis

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

GeneLLM: A Large cfRNA Language Model for Cancer Screening from Raw Reads

Contact Info

Product

Resources

About