analysis/pedigrees/Simulations/README.md

In ScribnerLab/SeaLampreyRapture:

There are there major steps when analyzing a COLONY BestConfig (BC) file generated using simulated offspring genotypes where the user provides no parental genotypes. When COLONY does not have candidate parental genotypes it will infer a parent, and its associated genotypes. These parents are given generic names (e.g., 1 or #3). Given that the user does not know which column represents each sex, each parent (P) column is given a generic column name (P1, P2), and the inferred parent IDs are changed slightly to reflect the header columns (e.g., 1 and #3 are changed to P1_1 and P2_3). It is also important to note that the offspring ID column (off) contains information about the known parents that produced each offspring, which is embedded in each offspring ID. For instance, the user knows that Mom 1 (m1) and Dad 1 (d1) produced offspring “o1_m1_d1”. Step 2 consists of evaluating all unique pairwise comparisons among all offspring genotyped (i.e., dyads). For each dyad (i.e. row) the user can determine the inferred relationship (IR) and known relationship (KR) between the two individuals by counting how many parents the two individuals share. That is, full-siblings (FS, 2 parents shared), half-siblings (HS, 1 parent shared), and unrelated (UR, 0 parents shared) dyads can be determined using both the Inferred and Known Parent IDs. Using information in step three the accuracy and false negative rate can be calculated for each simulation. See definitions and example in Figure for more details.

ScribnerLab/SeaLampreyRapture documentation built on Jan. 10, 2020, 9:18 a.m.