New algorithm facilitates personalized treatments for genetic disorders and cancer

The RTD detective computational model, developed by CRG researchers and published in Nature Genetics, predicts the efficacy of drugs for pathologies caused by genetic mutations. Based on data from more than 5000 patients, the system promises to accelerate the development of targeted therapies and optimize clinical trials.

A computational model developed by experts from the Institute for Research in Biomedicine (IRB) and the Centre for Genomic Regulation (CRG) in Barcelona can predict which drugs will be most effective in treating diseases caused by DNA mutations that lead to the synthesis of incomplete proteins.

The findings, published today in Nature Genetics, mark an important step in personalizing treatment by matching patients (based on their specific mutations) with the most promising drug. The predictive model, a public-use tool called RTDetective, can accelerate the design, development and effectiveness of clinical trials for a wide variety of genetic disorders and cancers.

The predictive model, a public-use tool called RTDetective, can accelerate the design, development and effectiveness of clinical trials for a wide variety of genetic disorders and cancers.

Incomplete proteins originate when protein synthesis suddenly stops. In our bodies, this is caused by the appearance of 'nonsense mutations' which act as a stop signal, causing the cellular machinery to come to a screeching halt. In many cases, these incomplete proteins are unable to carry out their function, leading to various disorders.

In fact, one in five diseases caused by mutations in a single gene is linked to these incomplete or unfinished proteins, including some types of cystic fibrosis and Duchenne muscular dystrophy. These premature stop signals also occur in tumor suppressor genes, whose function is to help control cell growth. Stop signals inactivate these genes and thus promote the development of cancer.

A comprehensive data analysis

Diseases caused by incomplete proteins can be treated with nonsense mutation suppression therapies, drugs that help cells ignore or 'read through' of the stop signals that appear during protein production. Cells with higher read rates will produce more full-length or nearly full-length proteins.

25% of diseases caused by mutations in a single gene are related to incomplete proteins

The study shows that, to date, clinical trials that work by reading through these stop signs, they have probably used ineffective drug-patient combinations. This is because the effectiveness of drugs depends not only on the mutation itself, but also on the genetic code surrounding it.

The researchers developed an experimental system based on human cell lines that allowed them to measure the efficacy of eight different drugs in 5800 premature stop signs disease-causing organisms. The data come from patient reports in freely accessible public archives such as ClinVar, as well as research projects such as The Cancer Genome Atlas (TCGA), who collected and analyzed genetic information from thousands of patients with cancer and genetic diseases, including premature stop mutations.

Accurate predictive models

To train the tool, they used effectiveness data for six of the eight drugs, across all possible stop signals in these 5800 patients. They found that a drug that works well to overcome one premature stop signal may not be effective for another, even within the same gene, due to the DNA sequence around the stop signal.

“Think of the DNA sequence as a highway, with a stop mutation appearing as a roadblock. We show that navigating through this obstacle depends heavily on the immediate surroundings. Some mutations are surrounded by well-marked detour routes while others are full of potholes or dead ends. This is what marks a drug’s ability to navigate obstacles and work effectively,” he explains. Ignasi Toledo, first author of the study.

At least one of the six drugs tested was predicted to achieve a 1% increase in reading in 87,3% of all possible stop signals, and a 2% increase in almost 40% of cases.

The researchers generated a large volume of data by testing many different combinations of drugs to read through the stop signs, resulting in a total of more than 140.000 individual measurementsThe vast amount of data generated allowed them to train accurate predictive models, which they used to create RTDetective.

The algorithm predicted the effectiveness of different drugs for each of the 32,7 million possible stop signals that can be generated in the human genome. At least one of the six drugs tested achieved a 1% increase in readthrough in 87.3% of all possible stop signals, and a 2% increase in nearly 40% of cases.

Personalized medicines

The results are promising because higher reading percentages generally correlate with better therapeutic outcomes. For example, Hurler's syndrome It is a severe genetic disorder caused by a nonsense mutation in the IDUA gene.

Previous studies have shown that, with Only 0,5% of “effective” reading, individuals can partially mitigate the severity of the disease by creating very small amounts of functional protein. RTDetective predicted that a reading above this threshold can be achieved with at least one of the drugs.

"With this system, the exact mutation is identified and then the computational model suggests which is the best drug for it."

Ben Lehner (ICREA)

“Imagine a patient is diagnosed with a genetic disorder. The exact mutation is identified through genetic testing and then a computational model suggests the best drug for them. This informed decision-making is the promise of personalized medicine that we hope to unlock in the future,” he explains. Ben Lehner, one of the lead authors of the study.

Treatment of tumors

The study also suggests how the new drugs can be rapidly assigned to the right patients. “When a new drug of this type is discovered, we can use this approach to quickly build a new predictive model and identify all the patients who are most likely to benefit,” adds Lehner.

Researchers plan to confirm the functionality of proteins produced by drugs that can read through stop signals, a crucial step to validate their clinical applicability. The team also plans to explore other strategies that can be used in combination with these therapies to increase the effectiveness of treatments, particularly in cancer.

“Our study not only opens new avenues for the treatment of hereditary genetic diseases, for which drugs that allow reading had already been tested, but also, and importantly, for the treatment of tumors, since most cancers have mutations that cause premature termination of proteins,” he concludes Fran Supek, co-lead author of the paper.

Reference: Toledano et al. «Genome-scale quantification and prediction of pathogenic stop codon readthrough by small molecules». Nature Genetics, 2024.

Source: SINC Agency

Source: CGR