Deep learning-based text mining for interpretation of omics data

Project Information

DeepTextNet

Grant agreement ID: 101023676

DOI

10.3030/101023676

Project closed

EC signature date 19 March 2021

Start date 1 November 2021

End date 31 October 2023

Funded under

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Total cost

€ 207 312,00

EU contribution

€ 207 312,00

Coordinated by

KOBENHAVNS UNIVERSITET
Denmark

Periodic Reporting for period 1 - DeepTextNet (Deep learning-based text mining for interpretation of omics data)

Reporting period: 2021-11-01 to 2023-10-31

Summary of the context and overall objectives of the project

In the rapidly evolving field of biomedical research, understanding the intricate network of molecular interactions is crucial. These interactions are the foundation of biological processes and diseases. However, deciphering this complex web has been a significant challenge due to the sheer volume and complexity of available data. This project aimed to tackle this challenge by harnessing the power of state-of-the-art text mining and network analysis techniques.

The insights gained from understanding molecular interactions have far-reaching implications for society, particularly in healthcare. By unraveling the details of these interactions, we can better comprehend disease mechanisms, discover potential drug targets, and develop more effective treatments. This project's outcomes are not just a leap forward for scientific understanding but also a step towards improving health outcomes and the quality of life.

The project had two primary goals. The first was to develop a next-generation text-mining technology using deep neural networks to extract molecular interactions from biomedical literature. This technology aims to transform how we gather and interpret complex biological data, making it more efficient and comprehensive. The second goal was to create advanced molecular networks and develop a novel methodology for integrated network analysis on large-scale omics data. This approach was intended to provide a deeper understanding of the molecular interactions and their implications in biological systems, thereby enhancing our ability to analyze and interpret vast amounts of omics data. Throughout the project, I focused on achieving these objectives while overcoming various challenges and adapting to unforeseen circumstances. The results have not only advanced the state-of-the-art in bioinformatics but also set the stage for future innovations in understanding and utilizing molecular interaction networks.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

From the outset, this project embarked on an ambitious journey to expand how we understand molecular interactions in biomedical research. The work was divided into several key areas, each contributing to the overall success of the project.

A significant part of the project involved developing a cutting-edge text mining system. This system uses advanced deep-learning techniques to extract information about molecular interactions from vast amounts of biomedical literature. This achievement marks a significant leap in our ability to process and understand biological data efficiently.
Another crucial aspect was the creation of comprehensive molecular networks. These networks integrate the information extracted from the text-mining process and apply it to real-world biological data. The result is a detailed map of molecular interactions, offering deeper insights into the biological processes and their implications for health and disease.

The outcomes of this project have been widely shared and utilized in various ways:

1. Enhancing the STRING database: The project's findings have significantly improved the STRING database, a critical resource for the scientific community. Continuous enhancement of the database allows for a more comprehensive and valuable resource for researchers worldwide.
2. Publications and presentations: The results have been disseminated through publications in open-access journals and presentations at international conferences. This widespread sharing has raised awareness and fostered collaboration within the scientific community.
3. Educational outreach: I have been involved in organizing seminars and conferences, sharing knowledge and skills with other researchers. Additionally, outreach activities have been conducted to engage the general public, enhancing the understanding and appreciation of science, particularly in artificial intelligence and bioinformatics.

The project has not only achieved its objectives but has also laid the groundwork for future advancements in the field. The methodologies and technologies developed are poised to have a lasting impact on how we analyze and interpret biological data, opening new avenues for research and healthcare innovations.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

This project has significantly advanced the field of bioinformatics, particularly in text mining and molecular network analysis. By leveraging deep learning techniques, the project has moved beyond traditional methods, enabling more efficient and accurate extraction of complex molecular interaction data from biomedical literature. This advancement not only enhances our understanding of biological processes but also paves the way for future progress in the field.

Specifically, the improvements made to the STRING database are expected to evolve further, offering even more detailed and comprehensive molecular interaction data to researchers globally. In addition, the methodologies and tools developed are anticipated to accelerate research in various biomedical fields, leading to new discoveries and insights into complex diseases. And finally, the open-source nature of the text mining system paves the way for its application in diverse research areas.

There are also several potential socio-economic impacts, as the project's outcomes are poised to contribute to the development of new diagnostic tools and treatments, potentially reducing healthcare costs and improving patient outcomes. Moreover, collaboration with industrial partners and the transfer of knowledge to commercial applications can stimulate innovation and economic growth in the biotechnology sector.

Perhaps more importantly, the project also has wider societal implications, as it has set a precedent for integrating advanced computational methods in bioinformatics education, preparing the next generation of researchers. It has also significantly raised awareness about the importance of bioinformatics and artificial intelligence in modern science, through public outreach activities, inspiring potential future scientists, and informing the public about scientific advancements.

In conclusion, the project's contributions to network biology and data integration are expected to have a long-lasting influence on how biological data is analyzed and interpreted, leading to breakthroughs in understanding human health and disease.

Marie curie staring at an image of a STRING network