Loading...
Please wait, while we are loading the content...
Similar Documents
Pedigree - a software package for computing inbreeding coefficients in large populations.
| Content Provider | Semantic Scholar |
|---|---|
| Author | Sargolzaei, Mehdi |
| Copyright Year | 2002 |
| Abstract | COMPUTING METHOD The genetic analysis of livestock records involves the inverse of additive relationship matrix (A). Computing inbreeding coefficients is necessary for direct forming the inverse of additive relationship matrix, which is used in mixed model equation. Henderson (1976) and Quaas (1976) implicitly presented methods for computing inbreeding coefficients. Henderson’s method requires storage of a large matrix. In the method of Quaas, memory requirement is linear with number of animals (N) and computation times is proportional to N, therefore computing time is limited with N. Golden et al (1991) proposed an algorithm based on Quaas’ method using sparse programming technique. Tier (1990) presented a fast algorithm for computing inbreeding coefficients in large population. The principle of Tier’s algorithm is based on this fact that we do not need to form complete A matrix and only a subset of this matrix is sufficient. Tier’s algorithm first determines which elements in the A matrix are required and then computes them. The memory requirement is about 0.9% as we must have in computing complete A matrix. In Tier’s algorithm for very large populations, we would have memory restrictions and if the computer forced to use hard disk memory then we would have significant decrease in speed. Meuwissen and Luo (1992) presented a fast method for computing inbreeding coefficients in large population, and one of its characteristic is high computational speed with low memory requirement. If the number of generations involved exceeds, say 12, the algorithm presented by Meuwissen and Luo becomes slow compared to Tier’s, because common ancestors are traced many times. In Pedigree software, two new algorithms have been used. The first algorithm (MS) has been used for increasing the speed of computing inbreeding coefficient, which first detects inbred animals without any computation, then with Meuwissen and Luo algorithm computes inbreeding coefficient for inbred animals only. The MS algorithm and the Meuwissen and Luo algorithm were compared with simulated data. When the number of inbred animals is lower than 35% of total population the MS algorithm will be faster than Meuwissen and Luo algorithm and it will be so faster when the number of animals and the number of generations are large (animals >50,000 and generations >10). The second algorithm (Pathway) has been used for determining paths, which relate parents of inbred animals to their common ancestors. The Pathway algorithm has lower speed than Tier and Meuwissen and Luo algorithms but provides more information for users. |
| Starting Page | 1 |
| Ending Page | 2 |
| Page Count | 2 |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.wcgalp.org/system/files/proceedings/2002/pedigree-software-package-computing-inbreeding-coefficients-large-populations.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
