MADET: a Manually Curated Knowledge Base for Microbiomic Effects on Efficacy and Toxicity of Anticancer Treatments

Zhang, Jie; Chen, Xiqian; Zou, Jiaxin; Li, Chen; Kang, Wanying; Guo, Yang; Liu, Sheng; Zhao, Wenjing; Mou, Xiangyu; Huang, Jiayuan; Ke, Jia

doi:10.1128/spectrum.02116-22

Cited by 3 publications

(1 citation statement)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Degradation products of red meat allow for the creation of a pro-inflammatory colonic microenvironment, and the gut microbiome plays a role in colorectal carcinogenesis. 31 , 32 In the present study, we found that more than 90% of the high-risk population and 45% of the low-risk population habitually consumed red meat, further supporting the association between red meat and CRC. The stool bacterial load is higher in individuals with high-grade dysplasia and CRC, and the value of fecal bacterial markers in CRC screening is gaining attention.…”

Section: Discussionsupporting

confidence: 78%

Optimization of screening strategies for colorectal cancer based on fecal DNA and occult blood testing

Yao

Sun

Xiong

et al. 2023

European Journal of Public Health

View full text Add to dashboard Cite

Background Fecal DNA and occult blood testing have been gradually developed for colorectal cancer (CRC) screening. Comparison of different testing strategies for these methods in CRC screening is in urgent need. This study aims to examine the efficacy of different testing strategies including multi-target fecal DNA testing, qualitative and quantitative fecal immunoassay tests (FITs). Methods Fecal samples were collected from patients diagnosed by colonoscopy. Tests using fecal DNA, quantitative FIT or qualitative FIT were performed on same fecal samples. Efficiency of different testing strategies within different populations was investigated. Results For high-risk populations (CRC and advanced adenoma), the positive rate of the three methods alone was 74.3–80%; the positive predictive values (PPVs) ranged from 37.3% to 77.8%, and the negative predictive values (NPVs) ranged from 86.3% to 92.2%. For combined testing strategies, the positive rate was 71.4–88.6%, PPVs ranged from 38.3% to 86.2%, and NPVs ranged from 89.6% to 92.9%. Parallel fecal multi-target DNA test and quantitative FIT appears to be superior when using a combined testing strategy. For the normal population, no significant difference was identified in efficacy between these methods when used alone and in combination. Conclusions Single testing strategy among the three methods is more suitable for the general population screening, and the combined testing strategy is more suitable for high-risk populations screening. The use of different combination strategies may have superiority in CRC high-risk population screening, but cannot conclude significant differences which may be attributed to the small sample size, large samples controlled trials are needed.

show abstract

Section: Discussionsupporting

confidence: 78%

Optimization of screening strategies for colorectal cancer based on fecal DNA and occult blood testing

Yao

Sun

Xiong

et al. 2023

European Journal of Public Health

View full text Add to dashboard Cite

show abstract

Databases and Tools for Microbial Genome and Human Microbiome Studies

Hojaisa

Jha

2023

Industrial Microbiology and Biotechnology

View full text Add to dashboard Cite

Leveraging pre-trained language models for mining microbiome-disease relationships

Karkera,

Acharya,

Palaniappan

2023

BMC Bioinformatics

View full text Add to dashboard Cite

Background The growing recognition of the microbiome’s impact on human health and well-being has prompted extensive research into discovering the links between microbiome dysbiosis and disease (healthy) states. However, this valuable information is scattered in unstructured form within biomedical literature. The structured extraction and qualification of microbe-disease interactions are important. In parallel, recent advancements in deep-learning-based natural language processing algorithms have revolutionized language-related tasks such as ours. This study aims to leverage state-of-the-art deep-learning language models to extract microbe-disease relationships from biomedical literature. Results In this study, we first evaluate multiple pre-trained large language models within a zero-shot or few-shot learning context. In this setting, the models performed poorly out of the box, emphasizing the need for domain-specific fine-tuning of these language models. Subsequently, we fine-tune multiple language models (specifically, GPT-3, BioGPT, BioMedLM, BERT, BioMegatron, PubMedBERT, BioClinicalBERT, and BioLinkBERT) using labeled training data and evaluate their performance. Our experimental results demonstrate the state-of-the-art performance of these fine-tuned models ( specifically GPT-3, BioMedLM, and BioLinkBERT), achieving an average F1 score, precision, and recall of over $$>0.8$$ > 0.8 compared to the previous best of 0.74. Conclusion Overall, this study establishes that pre-trained language models excel as transfer learners when fine-tuned with domain and problem-specific data, enabling them to achieve state-of-the-art results even with limited training data for extracting microbiome-disease interactions from scientific publications.

show abstract

MADET: a Manually Curated Knowledge Base for Microbiomic Effects on Efficacy and Toxicity of Anticancer Treatments

Abstract: Human microbiota plays an important role in mediating drug efficacy and toxicity in anticancer treatment. In this work, we developed a comprehensive online database, which documents over 480 microbiota-drug associations manually curated from 26 research articles.

Cited by 3 publications

References 27 publications

Optimization of screening strategies for colorectal cancer based on fecal DNA and occult blood testing

Optimization of screening strategies for colorectal cancer based on fecal DNA and occult blood testing

Databases and Tools for Microbial Genome and Human Microbiome Studies

Leveraging pre-trained language models for mining microbiome-disease relationships

Contact Info

Product

Resources

About