PARMESAN: A Cutting-Edge AI Tool Revolutionizing Genetic Disorder Treatments

Revolutionary AI Tool PARMESAN Accelerates Genetic Disorder Treatment Discovery

In the quest to unearth innovative treatments for genetic disorders, the scientific community faces a daunting challenge: the explosive growth of biomedical literature, often riddled with conflicting data. This proliferation of information has made it increasingly arduous for researchers to conduct comprehensive literature reviews, hindering progress in the field.

Enter Cole Deisseroth, a graduate student enrolled in the M.D./Ph.D. program, guided by the expertise of Drs. Huda Zoghbi and Zhandong Liu at the Jan and Duncan Neurological Research Institute (Duncan NRI), housed at Texas Children’s Hospital and Baylor College of Medicine. Cole spearheaded a groundbreaking study to develop a Natural Language Processing (NLP) tool known as PARsing ModifiErS via Article annotations, or PARMESAN for short.

PARMESAN, a game-changing AI-powered tool, is engineered to sift through the vast expanse of biomedical literature, gather up-to-date information, and predict potential drugs to rectify specific protein imbalances. A comprehensive overview of this tool and its remarkable capabilities has been published in The American Journal of Human Genetics.

Dr. Huda Zoghbi, distinguished service professor at Baylor College and the founding director of Duncan NRI, exclaims, “PARMESAN offers a wonderful opportunity for scientists to speed up the pace of their research and thus, accelerate drug discovery and development.”

What sets PARMESAN apart is its proficiency in harnessing curated data to forecast previously undiscovered associations.

Groundbreaking therapies for genetic disorders, scientists face a formidable challenge: the ever-expanding biomedical literature, often rife with conflicting information. To tackle this issue head-on, Cole Deisseroth, a graduate student in the M.D./Ph.D. program mentored by Drs. Huda Zoghbi and Zhandong Liu at the Jan and Duncan Neurological Research Institute (Duncan NRI) at Texas Children’s Hospital and Baylor College of Medicine, spearheaded a study leading to the creation of an innovative natural language processing (NLP) tool named PARsing ModifiErS via Article aNnotations (PARMESAN).

This cutting-edge tool performs a vital role by efficiently searching through public biomedical literature databases, such as PubMed and PubMed Central. PARMESAN not only identifies and ranks gene-gene and drug-gene regulatory relationships based on existing literature but stands out for its ability to predict potential novel drug-gene connections. Dr. Zhandong Liu, Chief of Computation Sciences at Texas Children’s Hospital and associate professor at Baylor College of Medicine, emphasizes that PARMESAN assigns an evidence-based score to each prediction, distinguishing it as a predictive powerhouse.

The unique feature of PARMESAN is that it not only identifies existing gene-gene or drug-gene interactions based on the available literature but also predicts putative novel drug-gene relationships by assigning an evidence-based score to each prediction.

The AI algorithms powering PARMESAN meticulously analyze studies outlining the intricate contributions within multistep genetic pathways. By assigning weighted numerical scores to reported interactions, the tool effectively prioritizes promising avenues for research. With over 18,000 target genes in its predictions and benchmarking studies indicating over 95% accuracy for the highest-scoring predictions, PARMESAN is a game-changer in expediting drug discovery and development.

“By pinpointing the most promising gene and drug interactions, this tool will allow researchers to identify the most promising drugs at a faster rate and with greater accuracy,” highlights Cole Deisseroth, underscoring the potential impact of PARMESAN on accelerating scientific advancements.

In a parallel effort at University College London, another groundbreaking application of AI is taking shape. A team of experts is leveraging artificial intelligence to enhance the palatability of medicines, particularly crucial for pediatric care where poor taste is a significant barrier to treatment adherence. Driven by an “electronic tongue,” an AI model is being developed to predict the taste of medicines by analyzing chemical descriptors.

This AI model, set to be an open-access tool, promises to revolutionize drug development by eliminating the need for early-stage taste trials. With a focus on bitterness and astringency, the main deterrents to medication adherence, this AI-driven approach aims to make drugs more palatable, especially for vulnerable populations such as children.

The AI model being developed will be an open-access tool, meaning that pharmaceutical development around the world can benefit from the data on the palatability of commonly used drugs.

As we witness these innovative applications of AI in the realms of genetic disorder research and drug palatability, it becomes evident that artificial intelligence is not just transforming how we approach these challenges but is doing so with an unprecedented level of efficiency and precision.

Dr. Zhandong Liu, Chief of Computation Sciences at Texas Children’s Hospital and associate professor at Baylor College of Medicine, emphasizes, “The unique feature of PARMESAN is that it not only identifies existing gene-gene or drug-gene interactions based on the available literature but also predicts putative novel drug-gene relationships by assigning an evidence-based score to each prediction.”

PARMESAN’s AI algorithms meticulously scrutinize studies that elucidate the roles of various players within complex genetic pathways. It then assigns a weighted numerical score to each reported interaction. Interactions consistently and frequently reported in the literature receive higher scores, while those with weak support or contradictions across studies are assigned lower scores.

Currently, PARMESAN provides predictions for more than 18,000 target genes, with benchmarking studies showcasing accuracy rates exceeding 95%.

In the words of Cole Deisseroth, “By pinpointing the most promising gene and drug interactions, this tool will allow researchers to identify the most promising drugs at a faster rate and with greater accuracy.”

As scientists strive to unlock the secrets of genetic disorders, PARMESAN stands as a beacon of hope, streamlining research processes, and propelling us closer to groundbreaking treatments.

In related news, scientists at University College London are harnessing artificial intelligence to address another pressing healthcare challenge: making medications more palatable, particularly for children who often struggle with the taste of drugs.

Medicines’ unappealing taste remains a significant obstacle, affecting medication adherence in various populations, including children, the elderly, and vulnerable groups. The bitter taste, in particular, poses challenges for patients taking long-term medications such as HIV antiretrovirals and tuberculosis antibiotics.

To overcome this hurdle, a team of experts at University College London has embarked on a mission to make drugs more palatable, faster than ever before, using an innovative approach. They employ an “electronic tongue” to collect data and create an AI model that predicts the taste of medicines.

This AI model dissects drugs into a series of chemical descriptors that determine taste and then maps them to predict bitterness levels. Dr. Hend Abdelhakim, an assistant professor at the UCL Global Business School for Health, explains, “We run a machine learning algorithm to basically try and see what’s the chemical structure, what’s the molecular structure, what are the other chemical-physical parameters that make it bitter, and try to see if there’s a relationship.”

Beyond bitterness, the AI can detect other taste qualities, including salty, sweet, sour, umami, and astringent. However, bitterness and astringency take center stage due to their impact on medication adherence.

This AI-driven approach promises to accelerate drug development by eliminating the need for early-stage human taste trials, a time-consuming and expensive process. Ultimately, it may not even require the electronic tongue.

The AI model being developed will be open-access, benefiting pharmaceutical development worldwide by providing data on the palatability of commonly used drugs. Taste is a crucial factor in medication adherence, particularly for long-term treatments like HIV and diabetes medications, making this research pivotal for healthcare.

Dr. Abdelhakim emphasizes, “It’s mainly a problem with children because they have a heightened sense of taste. Chronic medications, it does impact compliance. It’s not just the child being fussy.”

A study in the European Union revealed that 63% of children aged 10 to 18 identified “bad taste of medicines” as a barrier. Medication adherence is critical for conditions like HIV, where antiretroviral drugs often have an unfavorable taste.

Furthermore, ensuring adherence to antibiotics is essential in preventing antimicrobial resistance—a global health concern. Dr. Abdelhakim notes, “It’s actually a bigger problem for the rest of us. It might be one small cog in the machine, but if we can tackle it, then that’s good.”

In summary, these two groundbreaking endeavors—one involving PARMESAN’s AI-driven approach to genetic disorder treatment discovery and the other using AI to enhance medication palatability—highlight how cutting-edge technology is reshaping the landscape of medical research and healthcare, offering new hope and possibilities for patients worldwide.