Graph_Builder builds the graph from RNA-seq data. It reads the RNA-seq expression file in tab-delimited format and computes the pairwise pearson correlation matrix. The algorithm of Graph_Builder is a naive method which computes all-vs-all pearson correlation coefficients. The complexity of the algorithm is O(n^3). However, due to the MKL blas library, Graph_Builder runs very fast. In our tests, it took less than 1.5 hour to finish analysis of a RNA-seq data that contains ~180,000 transcripts and ~5,000 samples. The correlation matrix is stored in a triple column format and can be used as input of Markov Clustering Algorithm.

Make sure that you have the following installed

1. Python2.7 (Recommend Anaconda)

   1. numpy

   2. scipy

   3. Install dependency packages via pip:

      $ pip install scipy numpy

   4. Install dependency packages via conda:

      $ conda install scipy numpy

$ git clone https://github.com/Rinoahu/Graph_Builder

$python rna_seq/naive_cluster.py -i foo.tab -t .2 -o foo.xyz

-i: str. The name of a tab-delimited file. The 1st column stand for gene names/identifier and the rest columns stand for gene expression levels in different samples. For example:

#id	s1	s2	s3	...
g1	1.2	2.4	1.2 ...
g2	1.1	1.2	0 ...
g3	1.2	1.2	2.2 ...
...

-t: float. The threshold to filter low-correlation gene pairs. If the pearson correlation coefficient between gene i and j is less than the threshold, then, it will be set to 0. Default value is 0.2.

-o: str. The name of output file. The output is an 3 columns tab-delimited file where the 1st and 2nd columns stand for the gene identifiers, the 3rd column is the pearson's correlation coefficient.