A method and electronic device for prioritizing genes based on phenotype
A technology of prioritization and phenotyping, applied in the field of biological information, can solve the problems of low accuracy and low performance of WES data, and achieve the effect of good sorting effect.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0154] Validation of the performance evaluation of Example 1 in the expected collection of case report test sets
[0155] The inclusion criteria for case reports are: Inclusion criteria: 1) Case reports published in the following five scientific journals: American Journal of Medical Genetics Part A, BMC Medical Genetics, BMC Pediatrics, Frontiers in Genetics, Molecular Genetics & Genomic Medicine; 2) Case reports The report was published between November 25, 2018 and May 25, 2019; 3) the case report described the phenotype of at least one HPO record; 4) the case report identified the causative gene with the corresponding Entrez ID. The exclusion criteria for case reports were: 1) the case report identified multiple disease-causing genes in a single individual; 2) the case report described chromosomal abnormalities.
[0156]In this verification example, 243 case reports were expected to be collected first, 58 case reports were excluded according to the exclusion criteria, and t...
Embodiment 2
[0160] Verify the performance evaluation of Example 2 on the real patient test set
[0161] We investigated the performance of our method using real patient data, cohort 1 of the real patient test set consisted of 784 singleton patients diagnosed according to Baylor Genetics. We evaluate the performance of our method, Phenolyzer, Phenomizer, Phrank and HANRD on this real patient test set. Figure 11 A schematic diagram of the evaluation results of this verification example is shown, Figure 11 Including three graphs A, B and C, graph A shows the performance of this method, Phrank, Phenomizer, HANRD and Phenolyzer on the prioritization of disease-causing genes, as can be seen from graph A, in this evaluation, this method The probability of ranking the disease-causing gene in the Top 10 in the test set of real patients is 11.7%, and the probability of ranking the disease-causing gene in the Top 10 by Phrank, Phenomizer, HANRD and Phenolyzer are respectively: 9.6%, 5.2%, 7.1% an...
Embodiment 3
[0163] Verification Example 3 Increasing the weight of precise phenotypes can significantly improve the performance of this method in the comprehensive test set
[0164] Such as Figure 12 As shown, the linear regression between the number of input phenotypes and the grade of the disease-causing gene showed that there was no significant relationship between the number of phenotypes and the grade of the disease-causing gene (P = 0.513). However, the percentages of accurate and imprecise phenotypes were positively correlated with the detection performance of the software (P = 2.20e-16 and P = 3.79e-8, respectively). In contrast, the percentage of noisy phenotypes was negatively correlated with the detection performance of the software (P = 2.20e-16). This result strongly suggests that increasing the weight of precise or imprecise phenotypes can improve the performance of the method.
[0165] We first constructed a comprehensive test set containing different combinations of pre...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


