The Relationship Between Polygenic Risk Scores and Clinical Phenotype in Patients with Phenylketonuria: Genetic Prediction with the Random Forest Model

Yıl 2025, Cilt: 35 Sayı: 4, 728 - 735, 29.08.2025

Ebru Marzioğlu Özdemir , Ozkan Bagci

https://doi.org/10.54005/geneltip.1697947

Öz

Summary:
Background /Aims:
Pathogenic variations in the PAH gene cause phenylketonuria (PKU), a monogenic metabolic disorder. Individuals with the same mutation often exhibit phenotypic variability despite the monogenic nature of the condition. The aim of this study is to create a prediction model using the Random Forest (RF) machine learning algorithm and to examine the relationship between polygenic risk scores (PRS) and phenotypic severity in PKU patients.
Methods:
In this study, clinical exome sequencing data obtained from 174 PKU patients with molecular validation were retrospectively examined. Approximately 18,000 common variants were retained after being filtered by population frequency and quality for individual-level analysis. All eligible variants (excluding PAH mutations) were used to calculate PRS, and RF (1000 trees, maximum depth = 5) was used for modeling. International criteria were used to classify patients into mild, moderate, and severe phenotypes. Pearson correlation and ROC analysis were used to evaluate the model's performance.
Findings:
The RF-based PRS model had a high accuracy rate in predicting phenotypic severity (AUC = 0.91, overall accuracy = 84.3%). There was a significant correlation between PRS values and the severity of the phenotype (r = 0.68, p < 0.001). Severe clinical phenotypes were more common in patients with higher PRS. Variants in genes associated with phenylalanine metabolism (e.g., GCH1, QDPR, PTS) were the most significant contributors to risk prediction according to feature importance analysis results.
Results:
The results indicate that PRS modeling combined with machine learning could be a useful method for predicting the severity of phenotypes in monogenic disorders such as PKU. This integrative approach highlights the regulatory effect of a polygenic background and suggests that PRS could support clinical risk assessment and personalized treatment plans. However, before clinical application, it is very important to validate in various populations.

Anahtar Kelimeler

Phenylketonuria , polygenic risk score , genomic modeling , machine learning , phenotypic prediction

Fenilketonürili Hastalarda Poligenik Risk Skorlarının Klinik Fenotiple İlişkisi: Random Forest Modeli ile Genetik Öngörü

Öz

Özet:
Arka Plan / Amaçlar:
PAH genindeki patojenik varyasyonlar, monogenik bir metabolik bozukluk olan fenilketonüriye (PKU) neden olur. Aynı mutasyona sahip bireyler, durumun monogenik doğasına rağmen genellikle fenotipik değişkenlik gösterirler. Bu çalışmanın amacı, Random Forest (RF) makine öğrenimi algoritmasını kullanarak bir tahmin modeli oluşturmak ve PKU hastalarında poligenik risk skorları (PRS) ile fenotipik şiddet arasındaki ilişkiyi incelemektir.
Metod
Bu çalışmada, 174 PKU hastasından elde edilen klinik ekzom dizileme verileri retrospektif olarak moleküler doğrulama ile incelendi. Yaklaşık 18.000 yaygın varyant, bireysel düzeyde analiz için popülasyon frekansı ve kaliteye göre filtrelendikten sonra tüm uygun varyantlar (PAH mutasyonları hariç) PRS hesaplamak için kullanıldı ve modelleme için RF (1000 ağaç, maksimum derinlik = 5) kullanıldı. Hastaları hafif, orta ve şiddetli fenotiplere sınıflandırmak için uluslararası kriterler kullanıldı. Pearson korelasyon analizi ve ROC analizi, modelin performansını değerlendirmek için kullanıldı.
Bulgular:
RF tabanlı PRS modeli, fenotipik şiddetin tahmininde yüksek bir doğruluk oranına sahipti (AUC = 0.91, genel doğruluk = %84.3). PRS değerleri ile fenotipin şiddeti arasında anlamlı bir korelasyon vardı (r = 0.68, p < 0.001). Klinik fenotiplerin şiddetli olanları, daha yüksek bir PRS'ye sahip hastalarda daha yaygındı. Fenilalanin metabolizması ile ilişkili genlerdeki varyantlar (örneğin, GCH1, QDPR, PTS), özelliklerin önem analizi sonuçlarına göre risk tahminine en önemli katkıda bulunanlardı.
Sonuçlar:
Sonuçlar, PRS modellemesinin makine öğrenimi ile birleştirilmesinin, PKU gibi monogenik bozukluklarda fenotiplerin şiddetini tahmin etmek için yararlı bir yöntem olabileceğini göstermektedir. Bu bütünleştirici yaklaşım, çok genli bir arka planın düzenleyici etkisini vurgular ve PRS'nin klinik risk değerlendirmesi ve kişiselleştirilmiş tedavi planlarını destekleyebileceğini öne sürer. Ancak, klinik uygulamadan önce, çeşitli popülasyonlarda doğrulamak çok önemlidir.

Anahtar Kelimeler

Fenilketonüri , poligenik risk skoru , genomik modelleme , makine öğrenimi , fenotipik öngörü

Kaynakça

• 1. Berga-Švītiņa E, Miklaševičs E, Fischer K, Vilne B, Mägi R. Polygenic risk score predicts modified risk in BRCA1 pathogenic variant carriers in breast cancer patients. Cancers (Basel). 2023;15(11):2957. https://doi.org/10.3390/cancers15112957
• 2. Blau N, Longo N, van Spronsen FJ. PKU: Current management and future developments. Mol Genet Metab. 2021;132(1):1–12. https://doi.org/10.1016/j.ymgme.2021.04.003
• 3. Christoffersen M, Tybjærg‐Hansen A. Polygenic risk scores: How much do they add? Curr Opin Lipidol. 2021;32(3):157–62. https://doi.org/10.1097/mol.0000000000000759
• 4. De Vincentis A, Tavaglione F, Jamialahmadi O, et al. A polygenic risk score to refine risk stratification and prediction for severe liver disease by clinical fibrosis scores. Clin Gastroenterol Hepatol. 2022;20(3):658–73.e6. https://doi.org/10.1016/j.cgh.2021.05.056
• 5. Fahed AC, Wang M, Homburger JR, et al. Polygenic background modifies penetrance of monogenic variants for tier 1 genomic conditions. Nat Commun. 2020;11(1):3635. https://doi.org/10.1038/s41467-020-17374-3
• 6. Fang Y, Gao J, Guo Y, Li X, Yuan E, Zhang L. Allelic phenotype prediction of phenylketonuria based on the machine learning method. Hum Genomics. 2023;17(1). https://doi.org/10.1186/s40246-023-00481-9
• 7. Goodrich J, Singer‐Berk M, Son R, et al. Determinants of penetrance and variable expressivity in monogenic metabolic conditions across 77,184 exomes. Nat Commun. 2021;12(1). https://doi.org/10.1038/s41467-021-23556-4
• 8. Groenendyk JW, Ahmed ST, Fanidi A, et al. Implementation of population-based polygenic risk scores: A conference at the Pritchard Lab. BMC Med Genomics. 2021;14(1):91. https://doi.org/10.1186/s12920-021-00942-1
• 9. Groenendyk JW, Greenland P, Khan SS. Incremental value of polygenic risk scores in primary prevention of coronary heart disease. JAMA Intern Med. 2022;182(10):1082–9. https://doi.org/10.1001/jamainternmed.2022.3171
• 10. Honda S, Ikari K, Yano K, et al. Association of polygenic risk scores with radiographic progression in patients with rheumatoid arthritis. Arthritis Rheumatol. 2022;74(5):791–800. https://doi.org/10.1002/art.42051
• 11. Leal-Witt M, Rojas-Agurto E, Muñoz-González M, et al. Risk of developing insulin resistance in adult subjects with phenylketonuria: Machine learning model reveals an association with phenylalanine concentrations in dried blood spots. Metabolites. 2023;13(6):677. https://doi.org/10.3390/metabo13060677
• 12. Lewis CM, Vassos E. Polygenic risk scores: From research tools to clinical instruments. Genome Med. 2020;12(1):44. https://doi.org/10.1186/s13073-020-00742-5
• 13. Lu T, Forgetta V, Keller-Baruch J, et al. Improved prediction of fracture risk leveraging a genome-wide polygenic risk score. Genome Med. 2021;13(1):16. https://doi.org/10.1186/s13073-021-00838-6
• 14. Luckett A, Hawkes G, Green H, et al. Type 1 diabetes genetic risk contributes to phenotypic presentation in monogenic autoimmune diabetes. Diabetes. 2024;74(2):243–8. https://doi.org/10.2337/db24-0485
• 15. Ma W, Lau YL, Yang W, Wang YF. Random forests algorithm boosts genetic risk prediction of systemic lupus erythematosus. Front Genet. 2022;13:902793. https://doi.org/10.3389/fgene.2022.902793
• 16. Mullins N, Forstner AJ, O’Connell KS, et al. Genome-wide association study of more than 40,000 bipolar disorder cases provides new insights into the underlying biology. Nat Genet. 2021;53(6):817–29. https://doi.org/10.1038/s41588-021-00857-4
• 17. Page M, Vance ET, Cloward M, et al. The Polygenic Risk Score Knowledge Base: A centralized online repository for calculating and contextualizing polygenic risk scores. Res Sq [Preprint]. 2021. https://doi.org/10.21203/rs.3.rs-799235/v1
• 18. Pattee J, Pan W. Penalized regression and model selection methods for polygenic scores on summary statistics. PLoS Comput Biol. 2020;16(10):e1008271. https://doi.org/10.1371/journal.pcbi.1008271
• 19. Sipeky C, Talala K, Tammela TLJ, et al. Prostate cancer polygenic risk score and prediction of lethal prostate cancer. Sci Rep. 2020;10(1):17027. https://doi.org/10.1038/s41598-020-74172-z
• 20. Song Y, Yin Z, Zhang C, et al. Random forest classifier improving phenylketonuria screening performance in two Chinese populations. Front Mol Biosci. 2022;9:986556. https://doi.org/10.3389/fmolb.2022.986556
• 21. Timasheva Y, Balkhiyarova Z, Avzaletdinova DS, et al. Integrating common risk factors with polygenic scores improves the prediction of type 2 diabetes. Int J Mol Sci. 2023;24(2):984. https://doi.org/10.3390/ijms24020984
• 22. van Spronsen FJ, van Wegberg AMJ, Ahring K, et al. The ongoing challenge of phenylketonuria: Trends and developments. Nat Rev Endocrinol. 2017;13(7):405–18. https://doi.org/10.1038/nrendo.2017.51
• 23. Wang B, Irizar H, Thygesen JH, et al. Psychosis endophenotypes: A gene-set-specific polygenic risk score analysis. Schizophr Bull. 2023;49(6):1625–36. https://doi.org/10.1093/schbul/sbad088
• 24. Wei J, Shi Z, Na R, et al. Calibration of polygenic risk scores is required prior to clinical implementation: Results of three common cancers in UK Biobank. J Med Genet. 2022;59(3):243–7. https://doi.org/10.1136/jmedgenet-2020-107286
• 25. Wells QS, Bagheri M, Aday AW, et al. Polygenic risk score to identify subclinical coronary heart disease risk in young adults. Circ Genom Precis Med. 2021;14(5):e003341. https://doi.org/10.1161/CIRCGEN.121.003341
• 26. Zekanowski C, Krajewska-Walasek M, Mierzewska H. Phenylketonuria: Molecular diagnostics and treatment perspectives. J Inherit Metab Dis. 2016;39(5):695–705. https://doi.org/10.1007/s10545-016-9953-9
• 27. Zhu Z, Gu J, Genchev G, et al. Improving the diagnosis of phenylketonuria by using a machine learning–based screening model of neonatal MRM data. Front Mol Biosci. 2020;7:115. https://doi.org/10.3389/fmolb.2020.00115