Clustering a two-D matrix

[chayan]

Hii All,

I have generated a 2-D matrix based on NGS data which has cog scores in one axis (Y-axis) and sample no in other axis. Now i want to cluster this matrix based on the cog scores to find similarity between different sampling points and finally construct a UPGMA tree based on the output. Can any one suggest me a simple way to do this?? I don't have much knowledge about this...

Thanks for any help in advance...

Regards

Chayan

[GenoMax]

Try MEGA: http://megasoftware.net/

You should read the manual first to see if this would be appropriate to do with the dataset you have.

[southan]

You can use the Mclust package.

(http://www.stat.washington.edu/mclust/)

Change the parameter modelNames to obtain a suitable model (e.g., EII).

[chayan]

thanks to both of you. As i am going through both the manuals, it will take some time as i am a hardcore biology guy, i want to clear one my query.. my matrix structure is like this
A B C D E
i 20 30 15 43 87
ii 12 54 3 76 56
iii 45 78 4 99 54

now i want to cluster this based on the A, B, C, D, E based on the values of i, ii and iii...can i use "beta_diversity.py" from the qiime package..? then dont know basically how to generate to file format required for this as the prior steps are not same...


thanks

[chayan]

You can use the Mclust package.

(http://www.stat.washington.edu/mclust/)

Change the parameter modelNames to obtain a suitable model (e.g., EII).

thanks.. but i have never used R also... what i understand from the Mclust manual, if my matrix is like this

A B C D E
i 20 30 15 43 87
ii 12 54 3 76 56
iii 45 78 4 99 54

i should use the modelName "VII"/"VVV" or other unequal volume or varying volume from the multivariate mixture??

another thing, how to prepare the input file from the excel to feed into mclust??

thanks a lot