E-mail Updates Sign Up for E-mail Updates

How big data and AI can help you generate your scientific hypothesis

An Elsevier journal team works with Euretos to explore how machine learning and data analytics can guide research

© istock.com/liuzishan
© istock.com/liuzishan

Ask a researcher what challenges they face in their everyday work, and chances are they will tell you it’s about staying up to date on what’s happening in their field — keeping a close watch on what other research groups are doing and keeping an eye open for potential collaborators and new research opportunities.

So how much of that work can be handed over to software? A new pilot between Elsevier and the Euretos AI platform aims to use big data and machine learning to scan millions of journal articles and hundreds of databases to make connections and suggest new hypotheses for researchers to investigate.

Big-data technology

Staying up to date is all about making the right connections and figuring out the “must-read” articles. Researchers help each other with this. For example, they compile essential references or call out specific follow-up questions in their papers. But they can also benefit from increasingly powerful technology. Think about article recommendation engines that use sophisticated big-data algorithms to suggest which articles you should read.

But big data can do much more than fuel recommendation engines. It can become part of the scholarly discourse by performing a task it’s particularly well suited for: scanning through vast amounts of information to unearth connections that can lead to new scientific hypotheses. Data analytics can go deep into the article, identifying concepts and connections. When combined, these small bits of sometimes speculative or still amorphous knowledge can come together and materialize as a fundamental new insight that can guide researchers on what to research next.

“At Elsevier, we are excited to explore how cutting-edge technology can be combined with expert input from editors and other researchers to support the research dialogue," said Robbertjan Kalff, VP of Online Communities at Elsevier.

There are a lot of exciting developments in this space, including recent work by Elsevier technology teams (see, for example, Five reasons data is a key incredient for AI applications and Why today’s tech jobs need creative minds) as well as by tech start-ups and scale-ups like Netherlands-based Euretos and its AI Platform.

The 'Next Research' pilot

But even in the age of big data, it’s not only about the technology – it’s also about people. We need people to interpret the outcome of analyses, to scrutinize, to challenge, to agree or disagree, and to discuss. We need people to prevent “filter bubbles,” to help steer the conversation and connect people with new or unorthodox ideas.

This is where the scholarly journal comes to the foreground – playing its centuries-old role in organizing the scholarly discourse, managing an extensive network of domain experts who serve as editors or reviewers to make sure the journal’s readership stays up to date on critical developments in their field.

Inspired by recent technological advances, the publisher and editors of The International Journal of Biochemistry & Cell Biology (IJBCB) were keen to explore how big-data technology can help them in ways that are complementary to the input from this expert network. In particular, they wanted to see how big-data technologies could assist them not only with reporting on research but also to actively propel the research dialogue and guide the direction of next research.

Arie BaakTo put these ideas into practice, the journal team worked with other teams at Elsevier and with colleagues from Euretos and their AI platform to launch a pilot called “Next Research.” The project explores the power of combining big-data analytics with the human factor — aiming to find workflows in which humans and machines work together to accelerate the pace of research.

“The Euretos AI platform demonstrates the value of using big-data analytics to accelerate multi omics research,” said Euretos co-founder Arie Baak. “Making these capabilities part of hypothesis generation within the scientific discourse would raise its potential impact significantly.”

Here’s how the pilot works:

  • The publisher and editors select a recent top referenced IJBCB publication with an interesting open question in the discussion section.
  • This article is analyzed using the Euretos AI platform, bringing in additional knowledge from the vast body of literature and data, to turn the open question into specific new hypotheses that can be validated by future research.
  • The output of this analysis is collected and described in a discussion paper that is made available on the BioRN preprint server.
  • Researchers who have published in this field will be invited to discuss the paper in a dedicated Mendeley group called “Next Research” to help them find new research directions that are worth pursuing in their lab.
  • All relevant contributions to the discussion will be acknowledged in a follow-up paper to the original IJBCB publication.

Maddy Parsons, PhDDr. Maddy Parsons, Editor-in-Chief of IJBCB, wrote:

The Next Research discussion group is a fantastic new initiative that we’re very excited to be a part of at IJBCB. An open platform to discuss the ‘big picture’ questions emerging from different fields is a great way to get the global research community thinking collectively about important themes and new directions. It also offers an opportunity for exchanging ideas across disciplines and opening the door to new innovative ways to tackle key emerging problems. We encourage researchers at every level to get involved in the discussion and hopefully some exciting ideas will emerge!

We want to hear from you

The first discussion paper has just been shared in the Mendeley “Next Research” group. This is a pilot, which means that we will learn as we go and can iterate on the format between the various publications depending on results and feedback.

Being a pilot also means that we are eager to hear feedback, so – bot or not – feel free to leave your thoughts in the comment section below!

About Euretos

Euretos, founded in 2012 in the Netherlands, provides an AI platform mainly used by pre-clinical researchers for in-silico discovery and validation of targets and biomarkers. World-leading pharma, biotech and academic institutions use it to accelerate their multi omics research. By integrating over 200 public databases, the platform provides the largest single environment in which the latest multi omics data is interlinked to literature, experimental and clinical evidence. Euretos puts the power of bioinformatics in the hands of the researcher via an easy-to-use search engine and embedded analytics functions. This enables researchers to discover and evaluate how molecular mechanisms influence cell and tissue functions, and in turn mediate phenotypes and disease pathology.

Quick question for you

Which terms do you most associate with Elsevier? (check all that apply)

Data and analytics
Research platforms
Technology
Decision support tools
Publishing
Books and journals
Scientific articles
Healthcare content

Tags


Contributors


Comments


comments powered by Disqus