Machine learning approach for pooled DNA sample calibration

BMC Bioinformatics

Methodology Article

John M Henshall¹  Ashfaqur Rahman²  Andrew D Hellicar²  Daniel V Smith² 

[1] CSIRO Agriculture Flagship, Armidale, Australia;CSIRO Computational Informatics, Castray Esplanade, Hobart, Australia;

关键词: DNA pooling;    Calibration;    Machine learning;    SNP;   

DOI  :  10.1186/s12859-015-0593-1

 received in 2014-07-23, accepted in 2015-04-23,  发布年份 2015

来源: Springer

【 摘 要 】

BackgroundDespite ongoing reduction in genotyping costs, genomic studies involving large numbers of species with low economic value (such as Black Tiger prawns) remain cost prohibitive. In this scenario DNA pooling is an attractive option to reduce genotyping costs. However, genotyping of pooled samples comprising DNA from many individuals is challenging due to the presence of errors that exceed the allele frequency quantisation size and therefore cannot be simply corrected by clustering techniques. The solution to the calibration problem is a correction to the allele frequency to mitigate errors incurred in the measurement process. We highlight the limitations of the existing calibration solutions such as the fact they impose assumptions on the variation between allele frequencies 0, 0.5, and 1.0, and address a limited set of error types. We propose a novel machine learning method to address the limitations identified.ResultsThe approach is tested on SNPs genotyped with the Sequenom iPLEX platform and compared to existing state of the art calibration methods. The new method is capable of reducing the mean square error in allele frequency to half that achievable with existing approaches. Furthermore for the first time we demonstrate the importance of carefully considering the choice of training data when using calibration approaches built from pooled data.ConclusionThis paper demonstrates that improvements in pooled allele frequency estimates result if the genotyping platform is characterised at allele frequencies other than the homozygous and heterozygous cases. Techniques capable of incorporating such information are described along with aspects of implementation.

【 授权许可】

CC BY   
© Hellicar et al. 2015