View in EDS

A deep learning approach to prediction of blood group antigens from genomic data

Saved in:
Bibliographic Details
Title: A deep learning approach to prediction of blood group antigens from genomic data
Authors: Camous Moslemi, Susanne Sækmose, Rune Larsen, Thorsten Brodersen, Jakob T. Bay, Maria Didriksen, Kaspar R. Nielsen, Mie T. Bruun, Joseph Dowsett, Khoa M. Dinh, Christina Mikkelsen, Kati Hyvärinen, Jarmo Ritari, Jukka Partanen, Henrik Ullum, Christian Erikstrup, Sisse R. Ostrowski, Martin L. Olsson, Ole B. Pedersen
Source: Moslemi, C, Sækmose, S, Larsen, R, Brodersen, T, Bay, J T, Didriksen, M, Nielsen, K R, Bruun, M T, Dowsett, J, Dinh, K M, Mikkelsen, C, Hyvärinen, K, Ritari, J, Partanen, J, Ullum, H, Erikstrup, C, Ostrowski, S R, Olsson, M L & Pedersen, O B 2024, 'A deep learning approach to prediction of blood group antigens from genomic data', Transfusion, vol. 64, no. 11, pp. 2179-2195. https://doi.org/10.1111/trf.18013
Moslemi, C, Sækmose, S, Larsen, R, Brodersen, T, Bay, J T, Didriksen, M, Nielsen, K R, Bruun, M T, Dowsett, J, Dinh, K M, Mikkelsen, C, Hyvärinen, K, Ritari, J, Partanen, J, Ullum, H, Erikstrup, C, Ostrowski, S R, Olsson, M L & Pedersen, O B 2024, ' A deep learning approach to prediction of blood group antigens from genomic data ', Transfusion, vol. 64, no. 11, pp. 2179-2195 . https://doi.org/10.1111/trf.18013
Publisher Information: Wiley, 2024.
Publication Year: 2024
Subject Terms: Male, Neural Networks, Genotype, Denmark, convolutional neural network, deep learning, Blood Donors, Genomics, Blood Group Antigens/genetics, Genomics/methods, Computer, genetic prediction, Deep Learning, AI, denoising autoencoder, Blood Group Antigens, blood types, Humans, Female, Neural Networks, Computer, Illumina GSA, Finland, blood antigen
Description: BackgroundDeep learning methods are revolutionizing natural science. In this study, we aim to apply such techniques to develop blood type prediction models based on cheap to analyze and easily scalable screening array genotyping platforms.MethodsCombining existing blood types from blood banks and imputed screening array genotypes for ~111,000 Danish and 1168 Finnish blood donors, we used deep learning techniques to train and validate blood type prediction models for 36 antigens in 15 blood group systems. To account for missing genotypes a denoising autoencoder initial step was utilized, followed by a convolutional neural network blood type classifier.ResultsTwo thirds of the trained blood type prediction models demonstrated an F1‐accuracy above 99%. Models for antigens with low or high frequencies like, for example, Cw, low training cohorts like, for example, Cob, or very complicated genetic underpinning like, for example, RhD, proved to be more challenging for high accuracy (>99%) DL modeling. However, in the Danish cohort only 4 out of 36 models (Cob, Cw, D‐weak, Kpa) failed to achieve a prediction F1‐accuracy above 97%. This high predictive performance was replicated in the Finnish cohort.DiscussionHigh accuracy in a variety of blood groups proves viability of deep learning‐based blood type prediction using array chip genotypes, even in blood groups with nontrivial genetic underpinnings. These techniques are suitable for aiding in identifying blood donors with rare blood types by greatly narrowing down the potential pool of candidate donors before clinical grade confirmation.
Document Type: Article
File Description: application/pdf
Language: English
ISSN: 1537-2995
0041-1132
DOI: 10.1111/trf.18013
Access URL: https://pubmed.ncbi.nlm.nih.gov/39268576
https://vbn.aau.dk/da/publications/4f9f608b-0303-44a0-81ff-d5b29ee88a03
http://www.scopus.com/inward/record.url?scp=85203695657&partnerID=8YFLogxK
https://doi.org/10.1111/trf.18013
https://vbn.aau.dk/ws/files/754336085/Moslemi_et_al._2024_._A_deep_learning_approach_to_prediction_of_blood_group_antigens_from_genomic_data.pdf
https://curis.ku.dk/ws/files/413547537/Transfusion_2024_Moslemi_A_deep_learning_approach_to_prediction_of_blood_group_antigens_from_genomic_data.pdf
https://portal.findresearcher.sdu.dk/da/publications/dfe236f4-64ee-47aa-9884-5bb074600aec
https://doi.org/10.1111/trf.18013
https://pure.au.dk/portal/en/publications/50522a1c-a772-43cf-ba37-273fe15e2bf0
http://www.scopus.com/inward/record.url?scp=85203695657&partnerID=8YFLogxK
https://doi.org/10.1111/trf.18013
Rights: CC BY NC
Accession Number: edsair.doi.dedup.....ed08662a5443650eeb935b94449392f3
Database: OpenAIRE
Description
Abstract:BackgroundDeep learning methods are revolutionizing natural science. In this study, we aim to apply such techniques to develop blood type prediction models based on cheap to analyze and easily scalable screening array genotyping platforms.MethodsCombining existing blood types from blood banks and imputed screening array genotypes for ~111,000 Danish and 1168 Finnish blood donors, we used deep learning techniques to train and validate blood type prediction models for 36 antigens in 15 blood group systems. To account for missing genotypes a denoising autoencoder initial step was utilized, followed by a convolutional neural network blood type classifier.ResultsTwo thirds of the trained blood type prediction models demonstrated an F1‐accuracy above 99%. Models for antigens with low or high frequencies like, for example, Cw, low training cohorts like, for example, Cob, or very complicated genetic underpinning like, for example, RhD, proved to be more challenging for high accuracy (>99%) DL modeling. However, in the Danish cohort only 4 out of 36 models (Cob, Cw, D‐weak, Kpa) failed to achieve a prediction F1‐accuracy above 97%. This high predictive performance was replicated in the Finnish cohort.DiscussionHigh accuracy in a variety of blood groups proves viability of deep learning‐based blood type prediction using array chip genotypes, even in blood groups with nontrivial genetic underpinnings. These techniques are suitable for aiding in identifying blood donors with rare blood types by greatly narrowing down the potential pool of candidate donors before clinical grade confirmation.
ISSN:15372995
00411132
DOI:10.1111/trf.18013