Oops, you're using an old version of your browser so some of the features on this page may not be displaying properly.

MINIMAL Requirements: Google Chrome 24+Mozilla Firefox 20+Internet Explorer 11Opera 15–18Apple Safari 7SeaMonkey 2.15-2.23

e-Posters

49P - Identifying and validating networks of oncology biomarkers mined from the scientific literature

Date

06 Oct 2021

Session

e-Posters

Presenters

Kim Wager

Citation

Annals of Oncology (2021) 32 (suppl_6): S1345-S1371. 10.1016/annonc/annonc740

Authors

K. Wager1, D. Chari2, S. Ho2, T. Rees3, O. Penner4, B.J. Schijvenaars4

Author affiliations

  • 1 Oxford PharmaGenesis Ltd, Oxford/GB
  • 2 Pfizer Inc., 10017 - New York/US
  • 3 Oxford PharmaGenesis Ltd, OX13 5QJ - Oxford/GB
  • 4 Digital Science, London/GB
More

Abstract 49P

Background

Biomarkers are indicators of a biological state and have a pivotal role in estimations of disease risk, early detection, assessment of progression and outcomes prediction. Studies of cancer biomarkers are published daily, some are well characterized, while others are of growing interest. Managing this flow of information is challenging for scientists and clinicians. We sought to validate an AI-driven literature interrogation method and a publicly available tool to help researchers identify these emergent biomarkers.

Methods

Using the AI-based Dimensions literature platform, publications were identified through searches for biomarkers in proximity to terms relating to six cancer types. To highlight potential mechanistic insights, co-occurrences of one biomarker with another were then sought. To focus on biomarkers of emerging research interest, those with < 5 or > 1000 unique publication mentions were excluded. Network analysis was performed on the biomarker pairs and highly connected clusters determined. The mean publication growth rate for each cluster was calculated. Biological context for biomarker co-occurrence was validated manually.

Results

726 biomarkers were analyzed and 86 clusters identified across all cancer types. Based on publication growth rate, a renal cancer cluster of 16 biomarkers, with 47 biomarker pairs was selected for validation. The most commonly co-occurring pair was CXCL5–CXCL2. Contextual analysis showed all 34 mentioning publications to be valid associations, with 16 being renal cancer specific. Highlighted processes were mapped to the NCIT ontology and were consistent with a pro-inflammatory role for these chemokines on neutrophils in the tumor microenvironment, influencing angiogenesis, myeloid cell infiltration and metastasis. Evaluation of remaining biomarker pairs reveals that chemokines, matrix metalloproteinases and other regulators of cell-matrix interactions dominate. Future work will aim to improve disease-specific precision.

Conclusions

Our search method effectively finds relevant literature that could be missed with keyword searches, even where full text is available, and enables users to extract relevant biological insights.

Legal entity responsible for the study

Pfizer Inc.

Funding

Pfizer Inc.

Disclosure

K. Wager, T. Rees: Financial Interests, Personal, Full or part-time Employment: Oxford PharmaGenesis. D. Chari, S. Ho: Financial Interests, Personal, Full or part-time Employment: Pfizer Inc. O. Penner, B.J. Schijvenaars: Financial Interests, Personal, Full or part-time Employment: Digital Science.

This site uses cookies. Some of these cookies are essential, while others help us improve your experience by providing insights into how the site is being used.

For more detailed information on the cookies we use, please check our Privacy Policy.

Customise settings
  • Necessary cookies enable core functionality. The website cannot function properly without these cookies, and you can only disable them by changing your browser preferences.