Limits...
Haplotyping via minimum recombinant paradigm.

Hernández-Sánchez J, Knott S - BMC Proc (2009)

Bottom Line: A partial haplotyping algorithm consisting of four deterministic rules was almost as effective as a full one consisting of six deterministic rules, and took up to 5 times less time to compute.Haplotyping via the minimum recombinant paradigm is consistently reliable and computationally efficient.A single simulation is enough to produce a population-wide uncertainty estimate associated with a set of all reconstructed haplotypes.

View Article: PubMed Central - HTML - PubMed

Affiliation: Institute of Evolutionary Biology, University of Edinburgh, King's Buildings (Ashworth Laboratories), West Main Roads, EH9 3JT, Edinburgh, UK. jules.hernandez@ed.ac.uk

ABSTRACT

Background: Haplotypes can increase the power of gene detection over genotypes and are essential to estimate linkage disequilibrium.

Methods: Haplotyping was based on the minimum recombinant paradigm, whereby a phase is obtained only if it uniquely minimises the number of recombinants within a full sib family. Performance of this method was tested across three different data sets, consisting of genotypes and pedigree.

Results: The percentage of phased alleles ranged from ~80% to ~95%, and the percentage of correct phases reached ~99% in all cases. A measure of uncertainty was obtained via simulations. A partial haplotyping algorithm consisting of four deterministic rules was almost as effective as a full one consisting of six deterministic rules, and took up to 5 times less time to compute.

Conclusion: Haplotyping via the minimum recombinant paradigm is consistently reliable and computationally efficient. A single simulation is enough to produce a population-wide uncertainty estimate associated with a set of all reconstructed haplotypes.

No MeSH data available.


Related in: MedlinePlus

Pedigree structures. This figure shows the number of individuals (y-axes) against number of ancestors (x-axis), e.g. there are 359 sheep, 272 pigs and 239 individuals in the QTLMAS data with no ancestors (they are founders). The left y-axis is for sheep (clear bars) and pigs (grey bars), and the right y-axis for QTLMAS data (dark bars with top markers).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2654501&req=5

Figure 1: Pedigree structures. This figure shows the number of individuals (y-axes) against number of ancestors (x-axis), e.g. there are 359 sheep, 272 pigs and 239 individuals in the QTLMAS data with no ancestors (they are founders). The left y-axis is for sheep (clear bars) and pigs (grey bars), and the right y-axis for QTLMAS data (dark bars with top markers).

Mentions: Figure 1 shows the three pedigree structures. Pigs have a four generations pedigree, including founders. Eight individuals have 14 ancestors, the maximum found in this data set, which probably consisted of 2 parents, 4 grandparents and 8 great-grandparents. In contrast, the sheep pedigree was deeper with one individual having 32 ancestors, although most individuals (97%) had ≤14 ancestors. The simulated data set had the deepest and most complete pedigree with 7 generations including founders, and a maximum of ancestors for any individual, where i denotes generation.


Haplotyping via minimum recombinant paradigm.

Hernández-Sánchez J, Knott S - BMC Proc (2009)

Pedigree structures. This figure shows the number of individuals (y-axes) against number of ancestors (x-axis), e.g. there are 359 sheep, 272 pigs and 239 individuals in the QTLMAS data with no ancestors (they are founders). The left y-axis is for sheep (clear bars) and pigs (grey bars), and the right y-axis for QTLMAS data (dark bars with top markers).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2654501&req=5

Figure 1: Pedigree structures. This figure shows the number of individuals (y-axes) against number of ancestors (x-axis), e.g. there are 359 sheep, 272 pigs and 239 individuals in the QTLMAS data with no ancestors (they are founders). The left y-axis is for sheep (clear bars) and pigs (grey bars), and the right y-axis for QTLMAS data (dark bars with top markers).
Mentions: Figure 1 shows the three pedigree structures. Pigs have a four generations pedigree, including founders. Eight individuals have 14 ancestors, the maximum found in this data set, which probably consisted of 2 parents, 4 grandparents and 8 great-grandparents. In contrast, the sheep pedigree was deeper with one individual having 32 ancestors, although most individuals (97%) had ≤14 ancestors. The simulated data set had the deepest and most complete pedigree with 7 generations including founders, and a maximum of ancestors for any individual, where i denotes generation.

Bottom Line: A partial haplotyping algorithm consisting of four deterministic rules was almost as effective as a full one consisting of six deterministic rules, and took up to 5 times less time to compute.Haplotyping via the minimum recombinant paradigm is consistently reliable and computationally efficient.A single simulation is enough to produce a population-wide uncertainty estimate associated with a set of all reconstructed haplotypes.

View Article: PubMed Central - HTML - PubMed

Affiliation: Institute of Evolutionary Biology, University of Edinburgh, King's Buildings (Ashworth Laboratories), West Main Roads, EH9 3JT, Edinburgh, UK. jules.hernandez@ed.ac.uk

ABSTRACT

Background: Haplotypes can increase the power of gene detection over genotypes and are essential to estimate linkage disequilibrium.

Methods: Haplotyping was based on the minimum recombinant paradigm, whereby a phase is obtained only if it uniquely minimises the number of recombinants within a full sib family. Performance of this method was tested across three different data sets, consisting of genotypes and pedigree.

Results: The percentage of phased alleles ranged from ~80% to ~95%, and the percentage of correct phases reached ~99% in all cases. A measure of uncertainty was obtained via simulations. A partial haplotyping algorithm consisting of four deterministic rules was almost as effective as a full one consisting of six deterministic rules, and took up to 5 times less time to compute.

Conclusion: Haplotyping via the minimum recombinant paradigm is consistently reliable and computationally efficient. A single simulation is enough to produce a population-wide uncertainty estimate associated with a set of all reconstructed haplotypes.

No MeSH data available.


Related in: MedlinePlus