A Similarity Searching System for Biological Phenotype Images Using Deep Convolutional Encoder-decoder Architecture

摘要

Background: The BLAST (Basic Local Alignment Search Tool) algorithm has beenwidely used for sequence similarity searching. Analogously, the public phenotype images must beefficiently retrieved using biological images as queries and identify the phenotype with highsimilarity. Due to the accumulation of genotype-phenotype-mapping data, a system of searchingfor similar phenotypes is not available due to the bottleneck of image processing.Objective: In this study, we focus on the identification of similar query phenotypic images bysearching the biological phenotype database, including information about loss-of-function andgain-of-function.Methods: We propose a deep convolutional autoencoder architecture to segment the biologicalphenotypic images and develop a phenotype retrieval system to enable a better understanding ofgenotype–phenotype correlation.Results: This study shows how deep convolutional autoencoder architecture can be trained onimages from biological phenotypes to achieve state-of-the-art performance in a phenotypic imagesretrieval system.Conclusion: Taken together, the phenotype analysis system can provide further information on thecorrelation between genotype and phenotype. Additionally, it is obvious that the neural networkmodel of image segmentation and the phenotype retrieval system is equally suitable for anyspecies, which has enough phenotype images to train the neural network.