GPT Chat – which is a large language model (LLM) and belongs to the artificial intelligence (AI) family known as generative AI – has been a revelation to many people around the world, and, since its launch, has been used by professionals in many industries to help work-related tasks, such as editing and writing code. But could it also be used in the life sciences industry to advance drug discovery?
It’s no secret that AI has become a big deal in life sciences to advance drug discovery and development. This is especially true after the COVID-19 pandemic revealed AI to be an ideal tool to help find treatments and vaccines with greater speed and precision.
In fact, in a major breakthrough, it was recently reported that AI is rapidly helping discover a new antibiotic called abaucin, which can fight multi-drug-resistant bacteria.
While ChatGPT itself isn’t specifically designed for drug discovery, several companies are leveraging it as a useful tool to assist researchers in the process, which, overall, can be quite complex and anything that helps speed up the process is welcome. asset.
Assist in the drug discovery process
When we asked ChatGPT if it could be used to advance drug discovery, he responded saying that although it could used as a tool in the drug discovery process, ‘has limitations and is not a substitute for specialized software or expertise in the field.’
However, he said that it could help with drug discovery through data analysis, literature review, virtual screening, predictive modeling and decision support.
James Field, founder and chief executive officer (CEO) of LabGenius, also comments that ChatGPT is useful for querying data.
“Natural language processing (NLP) has been useful for querying data locked up in the scientific literature. An emerging solution is to query a large language model (LLM) including general models, such as ChatGPT, and domain-specific models, such as BioMedLM,” he said.
“For example, asking ChatGPT “What proteins can be targeted for the treatment of triple negative breast cancer with antibody therapy?” produce suggestions of EGFR, VEGF, PD-L1, PARP, and IGF-1R. While neither of these are revolutionary proposals, more domain-trained LLMs are likely to help accelerate target identification in the near future.
GPT-4 chat in particular appears to be very useful in aiding the drug discovery process, with OpenAI even explaining a number of possible ways to assist drug discovery in full. Technical reportpublished after the release of GPT-4.
Here, OpenAI states that GPT-4 could help find compounds similar to those researchers study, propose reengineering compounds and identify mutations that alter pathogenicity, and determine whether the compound is patentable.
Customizing ChatGPT for drug discovery
Additionally, ChatGPT is customizable in ways that allow researchers to work more easily with other forms of AI than their standard interfaces.
For example, an AI drug discovery company called Insilico Medicine has integrated ‘ChatPandaGPT’ into its PandaOmics platform, allowing researchers to have natural language conversations with its platform and efficiently navigate and analyze large data sets, in turn, more efficiently facilitating the discovery of potential therapeutic targets and biomarkers. .
ChatPandaGPT draws from a specialized knowledge base that enables it to provide accurate and detailed information related to molecular biology, therapeutic target discovery and pharmaceutical development.
By using natural language processing and machine learning algorithms, it can provide more personalized and relevant responses to researchers using the platform.
Use GPT Chat to develop biological experiments
The maker of the world’s first digital experiment platform for life sciences R&D, Synthace, also recently announced the integration of their platform with ChatGPT, to design protocols for biology experiments and automate lab work.
Experiments are difficult to design, plan, and automate in the lab, and the whole process can be time-consuming. However, Synthace’s ChatGPT prototype helped speed up the process, and allowed scientists to complete experiments in hours instead of weeks or more.
“With this prototype, Synthace uses ChatGPT to help scientists determine their experiments through natural language clues. When the scientist is ready, Synthace turns the experiment into instructions for the lab robot. ChatGPT has been trained in scientific literature, so it can interpret and design experiments, while Synthace is built to automate lab equipment,” explained Markus Gershater, co-founder and chief scientific officer (CSO) of Synthace.
When it comes to drug discovery, prototyping can actually help companies speed up this process, too.
“In our prototype, ChatGPT can be used to develop experiments that drug discovery companies need to optimize and use in their laboratories. Even companies that use AI to discover potential drugs still need to bring those candidates into the lab to run experiments on them. This is where Synthace comes into play,” said Gershater.
“I think the most exciting scientific discoveries will come from new proprietary data sets. This is because most scientific breakthroughs are often centered on new drug modalities, for which data are not available.”
James Field, founder and CEO of LabGenius
Main limitations of using GPT Chat for drug discovery
As much as ChatGPT may be for aiding drug discovery, it certainly has limitations, so much so that Stef van Grieken, co-founder and CEO of Cradle, said he would not currently recommend it for use in drug discovery. , for three reasons.
The first, he said, ChatGPT is often dishonest and will convincingly convey inaccurate information. Indeed, ChatGPT is known to produce ‘hallucinations’, in which the information it provides sounds plausible but is actually wrong or completely unrelated to the given context.
Van Grieken continues, saying the second reason is that ChatGPT finds it difficult to explain why it arrived at a certain answer or conclusion.
And, third: “ChatGPT has very limited access to relevant data and literature for drug discovery. Many scientific publishers do not allow access to their papers necessary to train these models, and many relevant experimental datasets and patents are likely lost.
However, van Grieken added that he thinks in the near future, companies will develop similar LLMs that are experts in understanding literature, data sets, and drug development patents and will be able to assist scientists by answering questions in real-time, finding relevant data, and summarizing the literature.
As for Field, ChatGPT’s limitations for drug discovery also stem from the fact that the data it provides is limited to what it can find on the internet.
“… I think the most exciting scientific discoveries will come from new proprietary data sets. This is because most scientific breakthroughs are often centered on new drug modalities, for which data are not available. However, thanks to recent advances in automation and disease modeling, companies can now create their own high-throughput, clinically relevant data,” he said.
Despite ChatGPT’s drawbacks, it appears to have a role to play in aiding drug discovery, though that role is currently limited. And, with OpenAI bringing GPT-4 soon after introducing the GPT-3.5 series to the world, who knows what the future holds for ChatGPT’s role for drug discovery if OpenAI decides to release a more advanced version.