R: Efficiently calculate double summation of product
1
0
Entering edit mode
7.0 years ago

Hello everyone,

I have a drug-target matrix A (5000 rows(target) * 6000 columns (drugs)), in which each element is a_ij, The second matrix is S, which is 6000 * 6000 matrix (drug-drug similarity matrix), where each element is s_lk. I would like to calculate the following formula: m_ij = sum(k=1..m, l=1..m) (a_il*a_jk*s_lk), [i.e., https://www.dropbox.com/s/5nw7ha5j8pkgcr9/sum.pdf?dl=0] which should a 5000 * 5000 matrix. Here m_ij will be used to calculate the target-target similarity as described in Bioinformatics Journal <<Drug-target interaction prediction through domain-tuned network-based inference>>

What is the efficient way to calculate such summation of product?

Can you give me a tip to calculate such formula, or some course to give a presentation to deal with such formula efficiently.

 

Thanks.

 

 

R matrices • 4.4k views
ADD COMMENT
1
Entering edit mode

If I am not wrong, the method described in the paper is available as an R package DT-Hybrid.

ADD REPLY
0
Entering edit mode

Thanks. Yes, I got it.

ADD REPLY
0
Entering edit mode

While we have many R experts here, I think this question is a better fit for stack overflow. If you were to provide some context on how this pertains to bioinformatics, this question might belong better here.

ADD REPLY
0
Entering edit mode

Thanks RamRS. I have revised my poster and added the bioinformatics information.

ADD REPLY
0
Entering edit mode

Thank you for adding details. I'm not great with R and your question needs some expert insight so I guess we'll have to wait for some of them to take a look at it :)

ADD REPLY
0
Entering edit mode

Thank you RamRS. I am also working on it. 

ADD REPLY
0
Entering edit mode

are either of the matrices sparse?

ADD REPLY
0
Entering edit mode

Matrice A is sparse and Matrice S is similarity matrix, which is not sparse.

ADD REPLY
0
Entering edit mode
7.0 years ago

This answer should be: A %*% S %*% t(A)

ADD COMMENT

Login before adding your answer.

Traffic: 3024 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6