Listen to this article
Scientists at Inventiva, a drug-discovery company, used to spend long hours collecting data manually from the machines that run their experiments so they could be analysed. Now, the business, based in the French city of Dijon, has deployed software that allows all its devices to talk to each other and share data across the business.
The technology gives the company a lead in the complex process of drug discovery, says Philippe Masson, head of screening and compound management at Inventiva, which is developing treatments for cancer and fibrosis. The new software “is significantly boosting the speed of identifying new compounds. It has made analysing the data 10 times faster”.
The previous process of copying and pasting data between systems was time-consuming and error-prone, says Mr Masson. “And the way in which data were presented made efficient analysis very difficult.” With drugs costing an average of $1.8bn to develop, anything that can accelerate matters and improve accuracy is hugely beneficial.
Despite advances in technology and understanding of biological systems, pharmaceuticals and biotech companies are often constrained by lack of capacity in process by the vast quantities of data generated from expensive research programmes. This is to the cost of both the company and potential patients who miss out on therapies.
Drug discovery laboratories have a longstanding need to integrate data from multiple disparate sources with different formats and much of which is high volume and complex, says Simon Kew, technology expert at PA Consulting.
“The data can range from the genomic profile of a patient to a simple measurement such as a drug’s solubility. Not many in the industry have fully integrated all their data on to one information management system,” Mr Kew says.
Inventiva’s 80 researchers work on internal programmes and in partnership with pharmaceutical companies including Chicago-based AbbVie. Their activities cover the entire drug development process, from “target validation”, identifying the particular protein or enzyme against which a drug should work, to selection of compounds for clinical trials.
The company has a library of 240,000 chemical compounds that it tests for their effectiveness against the target once it has been identified, a process known as high throughput screening (HTS).
Compounds from this screening process are tested in cells for efficacy. The aim is to reduce the potential for side-effects and increase potency and metabolic stability by looking at how the drug might be absorbed, distributed in the body and excreted. This involves taking pictures of the cells as they grow and analysing them, a process known as “high content screening” (HCS).
“HTS and HCS create huge quantities of data,” says Mr Masson. “We wanted a system that lets us quickly and securely analyse the results and display images speedily alongside the numbers.”
In early 2014, Inventiva began using a software package called ActivityBase from IDBS, a company located in Guildford, UK. The two companies have worked closely to deploy the software and create an integrated system to analyse data, including images of cell development, and store them securely.
“It is a very complex system,” Mr Masson says. “Although we bought off-the-shelf software, we had to customise it.” Results can now be seen in a more accessible form on screen, for example in graphs and curves.
“It’s particularly useful to have a more graphic visualisation of our data,” he adds. “This helps us evaluate the response at different [levels of] concentration and see whether a compound is highly active or not. Our researchers can now do more experiments, analyse more compounds and take decisions more quickly.”
The software also puts Inventiva in a much stronger position competitively because it speeds up internal processes and facilitates sharing data with clients. “Even if they are using different software, we can export data as Excel or text files, which makes communicating with other systems straightforward,” says Mr Masson.
Inventiva is now in discussions with IDBS about how the scope of the software could be extended, for example to improve the way the experimental control data are analysed.