Volume: 3, 2024
3rd International PhD Student’s Conference at the University of Life Sciences in Lublin, Poland:
ENVIRONMENT – PLANT – ANIMAL – PRODUCT
Abstract number: H020
DOI: https://doi.org/10.24326/ICDSUPL3.A026
Published online: 24 April 2024
ICDSUPL, 3, H020 (2024)
From genome to AI algorithm: novel perspectives in retinitis pigmentosa diagnosis and therapy
Agata Pietras-Baczewska3, Radosław Smagieł1*, Beata Gajda-Deryło3, Ewelina Cholewińska1, Przemysław Sołek1, Kamil Jonak2, Katarzyna Nowomiejska3, Robert Rejdak3, Katarzyna Ognik1
1 Department of Biochemistry and Toxicology, University of Life Sciences in Lublin, Akademicka 13, 20-950 Lublin, Poland
2 Department of Computer Science, Lublin University of Technology, Nadbystrzycka 38, 20-618 Lublin, Poland
3 Chair and Department of General and Paediatric Ophthalmology, Medical University of Lublin, Chmielna 1, 20-079 Lublin, Poland
* Corresponding author: radoslaw.smagiel@up.lublin.pl
Abstract
Genetic studies on retinitis pigmentosa (RP) are crucial for understanding the molecular basis of the disease. Due to its hereditary nature, identifying genetic markers can provide significant insights into the pathogenic mechanisms and lead to more effective diagnostic and therapeutic methods development. In the context of the prevalence of the Polish population, genetic studies on RP appear particularly important because this disease is classified as rare, necessitating early analysis of the population’s genetic profile. Furthermore, the dynamic advancement of modern DNA sequencing technologies presents new opportunities for precise genetic analysis of patients suspected of RP, paving the way for personalized therapeutic strategies and patient care standards improvements. The aim of the research is to seek correlations between genetic data and clinical observations in patients, considering changes monitored over a three-year period. Advanced artificial intelligence (AI) algorithms will be developed based on integrated data from genetic and clinical studies, which will be used for recognition and disease progression assessment. The presented studies are part of a larger project focused on genetic studies of retinitis pigmentosa in at least 200 patients. So far, whole exome sequencing (WES) analysis has been performed on 42 patient blood samples (CeGat, Germany), limited to identifying CNV changes of >=50 kb size and automatically selected SNV variants in disease-causing genes. Variants have been classified as pathogenic, likely pathogenic, or of uncertain clinical significance related to the patient’s clinical phenotype. The complete results of the studies for the first three patients have already revealed surprising data. In one case, coexistence of both a pathogenic variant (c.2296T>C; p.Cys766Arg) and a likely pathogenic variant (c.12268C>A; p.Pro40902Thr) in the USH2A gene, associated with RP, was detected. Additionally, in the same patient, a pathogenic variant of the RYR1 gene (1840C>T; p.Arg614Cys) was also identified, which is a significant finding associated with an increased risk of malignant hyperthermia type 1. In the remaining two cases, no variants were found in the analyzed regions, suggesting that the presence of a rare inherited disease is unlikely. Furthermore, none of these three cases showed the presence of variants of ≥50 kb, suggesting that the cause of retinitis pigmentosa may be more complex. Although mutational changes in DNA related to RP were identified in only one patient, the possibility of this condition occurring in the other two patients cannot be ruled out. There is a likelihood that additional genetic variants in introns, promoter regions, enhancer sequences, and long repeats may be responsible for disease development and were not considered in the analysis. There is also a risk of underestimating the occurrence of certain genetic variants, especially in regions with multiple genomic copies, and the lack of detection of low-frequency somatic mosaicism. It should also be noted that the classification of genetic variants may change in the future based on new scientific data, leading to a reassessment of disease risk and associated genetic variants. As a result, variants currently classified as benign, likely benign, or unidentified may prove crucial for retinitis pigmentosa pathogenesis understanding. Despite these limitations, pathogenic and likely pathogenic variants in RYR1 and USH2A genes, which are associated with RP but also with other conditions, have been detected. Therefore, further genetic and clinical studies are necessary for a fuller understanding of the pathogenesis of these diseases and personalized therapeutic strategies development. The use of artificial intelligence in RP identification will facilitate and expedite disease diagnosis and serve as a tool for progression assessment, enabling the selection of optimal and personalized therapy.
The research was carried out within the Lublin Digital Union (LUC) and was funded by the Ministry of Education and Science (actually Ministry of Science and Higher Education) in Poland, Grant No. MEiN/2023/DPI/2196.
Keywords: artificial intelligence, next-generation sequencing, mutation, retinitis pigmentosa
How to cite
A. Pietras-Baczewska, R. Smagieł, B. Gajda-Deryło, E. Cholewińska, P. Sołek, K. Jonak, K. Nowomiejska, R. Rejdak, K. Ognik, 2024. From genome to AI algorithm: novel perspectives in retinitis pigmentosa diagnosis and therapy. In: 3rd International PhD Student’s Conference at the University of Life Sciences in Lublin, Poland: Environment – Plant – Animal – Product. https://doi.org/10.24326/ICDSUPL3.A026