|
| ID | Methods | Advantages and disadvantages |
|
| 1 | Linkage analysis method genome-wide association analysis | Advantages: detecting the relationship between genetic loci and phenotypic trait |
| Disadvantages: low accuracy; time-consuming; high false positive rate; high cost |
|
| 2 | Expression Quantitative Trait Loci (eQTL) | Advantages: mining the relationship between genetic loci and quantitative trait |
| Disadvantages: high false positive rate; high cost |
|
| 3 | Mathematical and statistical methods | Advantages: high efficiency; low cost |
| Disadvantages: low accuracy; can not deal with the data with noise and nonlinear relationship; high computational complexity |
|
| 4 | Dimensionality Reduction Multifactor (MDR) | Advantages: detecting the epistasis interaction; low cost |
| Disadvantages: high calculation complexity |
|
| 5 | Clustering method | Advantages: establishing the global relationship; low cost |
| Disadvantages: cannot deal with data with nonlinear relationship; cannot detect the correlation between genes and phenotype |
|
| 6 | Support vector machine random forest method | Advantages: low cost; high efficiency |
| Disadvantages: the correctness depends on the quality of training set, but the training set is often hard to obtain |
|
| 7 | Bayesian method | Advantages: using the prior knowledge; realize the accurate calculation; low cost |
| Disadvantages: lack of network visibility; can not detect the epistasis interaction |
|
| 8 | Bayesian network | Advantages: processing data with noise and non-linear relationship; supporting different data types; high precision; low cost; detecting the epistasis interaction |
| Disadvantages: low learning efficiency; easy to cause local search |
|
