|
WISARD[wɪzərd] Workbench for Integrated Superfast Association study with Related Data |
|
This section describes about
WISARD reports missing rate of genotypes per variant and sample, and for family-based samples, it is also provided per families. Missing rate is sometimes called call rate and, for instance, individual call rate means the missing rate per individual. Large missing rate per variant or individuals can indicate genotyping errors, and they are often used for downstream QC. Summary of missing rate can be produced by adding --miss option as follows:
Running WISARD with --miss option produces the following results.
| sample.miss.lst is... | A report of genotype missingness per sample (TSV) | ||
| Column | Format | Modifier | Description | FID | string | NONE | Family ID of sample that genotype missingness investigated | IID | string | NONE | Individual ID of genotype missingness investigated | NMISS | integer | NONE | Number of missing genotypes | PROP | real | NONE | Proportion of missingess for the sample |
|---|---|---|---|
| family.miss.lst is... | A report of genotype missingness per family (TSV) | ||
| Detailed information about this format is not yet available | |||
| variant.miss.lst is... | A report of genotype missingness per variant (TSV) | ||
| Column | Format | Modifier | Description | CHR | real | NONE | Proportion of missingess for the sample | VARIANT | real | NONE | Proportion of missingess for the sample | POS | real | NONE | Proportion of missingess for the sample | ALT | real | NONE | Proportion of missingess for the sample | ANNOT | real | NONE | Proportion of missingess for the sample | NMISS | integer | NONE | Number of missing genotypes | PROP | integer | NONE | Proportion of missingess for the sample |
|---|---|---|---|
It filters sample(s) by calling rate(or genotyping rate).
For the dataset with dichotomous phenotype, WISARD can test whether the missing rates of variants are homogeneous between cases and controls with chi-square test. It can be conducted by adding --mistest option.
Running WISARD with --mistest option produces the following [prefix].mistest.res:
| mistest.res is... | A result of test of random missingness between case/control (TSV) | ||
| Column | Format | Modifier | Description | CHR | real | NONE | Overall Ts/Tv ratio in current chromosome | VARIANT | real | NONE | Overall Ts/Tv ratio in current chromosome | POS | real | NONE | Overall Ts/Tv ratio in current chromosome | ALT | real | NONE | Overall Ts/Tv ratio in current chromosome | ANNOT | real | NONE | Overall Ts/Tv ratio in current chromosome | P_MISTEST_(phenotype name) | real | --sampvar,--pname | p-value of random missingness test for given binary phenotype |
|---|---|---|---|
When there are multiple phenotypes in the dataset, all phenotypes must be dichotomous to perform missingness test. Note that an inclusion/exclusion of variants from dataset can be achieved with --mistest, by --incmistest or --filmistest. When there are multiple phenotypes in the dataset, a variant will be chosen or removed if at least one test has satisfied.