In microarray-based cancer classification, gene selection is an important issue owing to the large number of variables and small number of samples as well as its non-linearity. It is difficult to get satisfying results by using conventional linear statistical methods. Recursive feature elimination based on support vector machine (SVM RFE) is an effective algorithm for gene selection and cancer classification, which are integrated into a consistent framework. In this paper, we propose a new method to select parameters of the aforementioned algorithm implemented with Gaussian kernel SVMs as better alternatives to the common practice of selecting the apparently best parameters by using a genetic algorithm to search for a couple of optimal parameter. Fast implementation issues for this method are also discussed for pragmatic reasons. The proposed method was tested on two representative hereditary breast cancer and acute leukaemia datasets. The experimental results indicate that the proposed method performs well in selecting genes and achieves high classification accuracies with these genes.
Algorithms ; Breast Neoplasms ; genetics ; Female ; Genes ; Humans ; Leukemia, Myeloid, Acute ; genetics ; Models, Genetic ; Oligonucleotide Array Sequence Analysis ; methods ; Predictive Value of Tests

Objective: To explore the heterogeneity and correlation of clinical phenotypes and genotypes in children with disorders of sex development (DSD). Methods: A retrospective study of 1 235 patients with clinically proposed DSD in 36 pediatric medical institutions across the country from January 2017 to May 2021. After capturing 277 DSD-related candidate genes, second-generation sequencing was performed to analyzed the heterogeneity and correlation combined with clinical phenotypes. Results: Among 1 235 children with clinically proposed DSD, 980 were males and 255 were females of social gender at the time of initial diagnosis with the age ranged from 1 day of age to 17.92 years. A total of 443 children with pathogenic variants were detected through molecular genetic studies, with a positive detection rate of 35.9%. The most common clinical phenotypes were micropenis (455 cases), hypospadias (321 cases), and cryptorchidism (172 cases) and common mutations detected were in SRD5A2 gene (80 cases), AR gene (53 cases) and CYP21A2 gene (44 cases). Among them, the SRD5A2 mutation is the most common in children with simple micropenis and simple hypospadias, while the AMH mutation is the most common in children with simple cryptorchidism. Conclusions: The SRD5A2 mutation is the most common genetic variant in Chinese children with DSD, and micropenis, cryptorchidism, and hypospadias are the most common clinical phenotypes. Molecular diagnosis can provide clues about the biological basis of DSD, and can also guide clinicians to perform specific clinical examinations. Target sequence capture probes and next-generation sequencing technology can provide effective and economical genetic diagnosis for children with DSD.
3-Oxo-5-alpha-Steroid 4-Dehydrogenase/genetics* ; Child ; China/epidemiology* ; Cryptorchidism/genetics* ; Disorders of Sex Development/genetics* ; Female ; Genital Diseases, Male ; Genotype ; Humans ; Hypospadias/genetics* ; Male ; Membrane Proteins/genetics* ; Penis/abnormalities* ; Phenotype ; Retrospective Studies ; Steroid 21-Hydroxylase/genetics*