Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • News and Commentary
  • Published:

Merlin: faster linkage analysis with improved genotyping error detection

The Pharmacogenomics Journal volume 2, pages 139–140 (2002)Cite this article

Abecasis and colleagues report the development of a more efficient approach to linkage analysis.1 The approach is based on the Lander–Green algorithm in which each alternative gene flow pattern in a pedigree is considered separately.2 The first algorithm modification was to use binary trees instead of a bit indexed array, as implemented in Genehunter,3 to estimate the alternative patterns of gene flow through a pedigree. In the trees, each level in the tree represents a meiosis and each branch denotes transmission of a grandmaternal or grandpaternal allele. The trees allow efficient representation of alternative branches that are identical and equally likely.

Readers interested in more details about the algorithm will not only want to read the article in detail, but may examine the source code as well (provided in C++ at http://bioinformatics.well.ox.ac.uk/Merlin). For those who want to know more about how Merlin performs before trying it on their own data, simulations in the paper provide benchmark assessment for four different pedigree sizes. Simulations comparing Genehunter (v2.1), Allegro (v1.1b),4 and Merlin were performed on a 700 Mhz Pentium III with 2 gigabytes of RAM. In contrast to Merlin, Allegro required an additional 20 gigabytes of disk space for swapping of intermediate calculations. The advantage in terms of speed was not substantial for 1000 affected sibling pairs given 50 markers at 2 cM spacing. Allegro v 1.1b and Merlin did not differ substantially in speed for pedigrees in which twice the number of nonfounders − the number of founders (2n − f) was 25 and in which grandparents were typed in these three generation pedigrees. The exact analysis using Merlin failed to analyze pedigrees in which 2n − f was increased to 32. Therefore, sparse trees accounted for a small increase in speed and no increase in the size of pedigree that could be managed.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

References

  1. Abecasis GR et al . Nat Genet 2002 30: 97–101

  2. Lander ES, Green P . Proc Natl Acad Sci USA 1987 84: 2363–2367

  3. Kruglyak L et al . Am J Hum Genet 1996 58: 1347–1363

  4. Gudbjartsson DF et al . Nat Genet 2000 25: 12–13

  5. Abecasis GR et al . Am J Hum Genet 2001 68: 1463–1474

Download references

Author information

Authors and Affiliations

  1. Departments of Psychiatry and Pediatrics, University of Chicago, Chicago, IL, USA

    E H Cook Jr

Authors
  1. E H Cook Jr
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to E H Cook Jr.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Cook Jr, E. Merlin: faster linkage analysis with improved genotyping error detection. Pharmacogenomics J 2, 139–140 (2002). https://doi.org/10.1038/sj.tpj.6500102

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/sj.tpj.6500102

This article is cited by

Search

Advanced search

Quick links