Thanks to everyone for the great webcast yesterday. We had over 850 people register for the event and actually broke the record! Take that Bryce and Gabe!
If you would like to see the recording, view it at: Mixed Models: How to Effectively Account for Inbreeding and Population Structure in GWAS.
While preparing for this webcast, we chose to focus specifically on a few methods. But soon people in the community reached out and provided information on additional methods. (Thanks for sharing!) So here is a list of more mixed models if you want to dig in!
Methods We Covered:
EMMAX: Efficient Mixed-Model Association eXpedited (based on EMMA, which we didn’t cover)
Kang HM, Sul JH, Service SK, Zaitlen NA, Kong SY, Freimer NB, Sabatti C, Eskin E
http://genome.sph.umich.edu/wiki/EMMAX and http://mouse.cs.ucla.edu/emma/
GBLUP: Genomic Best Linear Unbiased Prediction
VanRaden PM
http://aipl.arsusda.gov/publish/jds/2008/abs91_4414.html
MLMM: Multi-locus mixed model
Segura V, Vilhjálmsson B, Platt A, Korte A, Seren U, Long Q, Nordborg, M
https://cynin.gmi.oeaw.ac.at/home/resources/mlmm
Other Methods We Didn’t Cover:
ASREML
Gilmour A
http://www.mmontap.org/resources/getting-started/asreml-r
Bayes-A,B,C and Cπ
Habier D, Fernando R, Kizilkaya K, and Garrick D
http://www.biomedcentral.com/1471-2105/12/186
Compressed MLM and P3D: Population Parameters Previously Determined
Zhang Z, Ersoz E, Lai CQ, Todhunter RJ, Tiwari HK, Gore MA, Bradbury PJ, Yu J, Arnett DK, Ordovas JM, and Buckler ES
http://www.nature.com/ng/journal/v42/n4/abs/ng.546.html
FaST-LMM: FActored Spectrally Transformed Linear Mixed Models
Listgarten J, Lippert C, Kadie CM, Davidson RI, Eskin E, and Heckerman D
http://research.microsoft.com/en-us/um/redmond/projects/mscompbio/Fastlmm/
GENMIX: Genealogy based Mixed Model
Sahana G, Mailund T, Lund MS, and Guldbrandtsen B
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0027061
GRAMMAR-Gamma (GenABEL R package)
Svishcheva G, Axenovich T, Belonogova N, van Duijn C, and Aulchenko Y
http://www.nature.com/ng/journal/v44/n10/full/ng.2410.html
http://hosho.ees.hokudai.ac.jp/~kubo/Rdoc/library/GenABEL/html/grammar.html
MASTOR: Mixed-model Association Score Test On Related individuals
Jakobsdottir J, McPeek MS
http://www.stat.uchicago.edu/~mcpeek/software/MASTOR/index.html
MixABEL package within GenABEL
Aulchenko YS, Ripke S, Isaacs A, and van Duijn CM
http://www.genabel.org/packages/MixABEL
MTMM: Multi-Trait Mixed Model
Korte A, Vilhjálmsson B, Segura V, Platt A, Long Q, and Nordborg M
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3432668/
QTLRel
Cheng R, Abney M, Palmer AA and Skol AD
http://cran.r-project.org/web/packages/QTLRel/
Unified Mixed Model in TASSEL
Yu J, Pressoir G, Briggs WH, Bi IV, Yamasaki M, Doebley JF, McMullen MD, Gaut BS, Nielsen MD, Holland JB, Kresovich S, and Buckler ES
http://www.maizegenetics.net/unified-mixed-model
What did I miss? I’m sure there are more out there – let me know in the comments!
Dear Greta Linse Peterson!
Could you answer me on a question. Is SVS7 package working with ILLUMINA files (for example, BovineSNP50)?
Sincerely, Michael
Dear Michael,
Thank you for your question! Yes, if you have your data in a Final Report Text File format you can import that into SVS (http://doc.goldenhelix.com/SVS/latest/importing_data.html#illumina). However, the preferred method is to install the Golden Helix GenomeStudio DSF Plug-in and export DSF files from GenomeStudio. (http://www.goldenhelix.com/Support/illumina_support.html). Note, if you only have iDat files, or do not have access to the GenomeStudio project you will need to contact the person that provided those and request the final reports or ask them to install the DSF plug-in and send you the DSF files. If you have any questions about this process please feel free to contact support@goldenhelix.com and we would be more than happy to help you import your Final Report Text Files or DSF files into an SVS Project.
Dear Greta Linse Peterson,
I have a question about the best step selection of MLMM model using SVS7 package. If there are several optimal steps according to different criteria, which step should i select?
Dear Jiaren Zhang,
Thank you for your question! The Segura 2012 paper recommends Extended BIC (Ext. BIC in SVS) or multiple-Bonferroni criterion (Bonferroni in SVS). They claimed that Ext. BIC was more stringent than Bonferroni. However, I found that it really depended on the model and the data itself, as sometimes Bonferroni was more stringent than Ext. BIC. I would look at Bonferroni, Mod. BIC and Ext. BIC and pick the first step that is optimal for any one of those criteria. I.e. the one that includes the fewest co-factors. If you had to pick just one criteria to use, I think the Ext. BIC is a good one to use, but I encourage you to read the Segura 2012 paper for their justification on these two criteria: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3386481/.