Skip to content

Commit

Permalink
Update predict_main.cpp
Browse files Browse the repository at this point in the history
  • Loading branch information
@WeiqiangZhou
WeiqiangZhou committed May 15, 2018
1 parent 2b7ecb5 commit b039ab3
Showing 1 changed file with 79 additions and 39 deletions.
118 changes: 79 additions & 39 deletions source/predict_main.cpp
View file
Expand Up @@ -6,42 +6,72 @@ int main(int argc, char *argv[])
char outfile[255];
char libfile[255]="./model_file.bin";
int write_flag = 0;
int locus_model = 0;
double up_bound = 14;
int opt;

if (argc == 2 && strcmp(argv[1], "-h") == 0)
if(argc == 1)
{
help_info();
return 1;
}
else if (argc == 5 && strcmp(argv[1], "-i") == 0 && strcmp(argv[3], "-o") == 0 )

while ((opt = getopt(argc,argv,"b:i:o:u:whl")) != EOF)
{
strcpy(infile, argv[2]);
strcpy(outfile, argv[4]);
switch(opt)
{
case 'b':
strcpy(libfile, optarg);
break;
case 'i':
strcpy(infile, optarg);
break;
case 'o':
strcpy(outfile, optarg);
break;
case 'u':
up_bound = atof(optarg);
break;
case 'w':
write_flag = 1;
break;
case 'h':
help_info();
return 1;
case 'l':
locus_model = 1;
break;
case '?':
std::cout << "Please input the correct parameters." << std::endl;
std::cout << "Standard output: BIRD_predict -b model_file.bin -i input_file.txt -o output_file.txt" << std::endl;
std::cout << "WIG output: BIRD_predict -b model_file.bin -i input_file.txt -o output_name -w" << std::endl;
return 1;
default:
abort();
}
}
else if (argc == 6 && strcmp(argv[1], "-i") == 0 && strcmp(argv[3], "-o") == 0 && strcmp(argv[5], "-w") == 0)

//output parameters
std::cout << "Using model file: " << libfile << std::endl;
std::cout << "Input file name: " << infile << std::endl;
std::cout << "Output file name: " << outfile << std::endl;
std::cout << "Using upper bound: " << up_bound << std::endl;
if(write_flag)
{
strcpy(infile, argv[2]);
strcpy(outfile, argv[4]);
write_flag = 1;
std::cout << "Output mode: wig file" << std::endl;
}
else if (argc == 7 && strcmp(argv[1], "-b") == 0 && strcmp(argv[3], "-i") == 0 && strcmp(argv[5], "-o") == 0)
else
{
strcpy(libfile, argv[2]);
strcpy(infile, argv[4]);
strcpy(outfile, argv[6]);
std::cout << "Output mode: data matrix" << std::endl;
}
else if (argc == 8 && strcmp(argv[1], "-b") == 0 && strcmp(argv[3], "-i") == 0 && strcmp(argv[5], "-o") == 0 && strcmp(argv[7], "-w") == 0)

if(locus_model)
{
strcpy(libfile, argv[2]);
strcpy(infile, argv[4]);
strcpy(outfile, argv[6]);
write_flag = 1;
std::cout << "Prediction mode: locus-level model" << std::endl;
}
else
{
std::cout << "Please input the correct parameters." << std::endl;
std::cout << "Standard output: BIRD_predict -b model_file.bin -i input_file.txt -o output_file.txt" << std::endl;
std::cout << "WIG output: BIRD_predict -b model_file.bin -i input_file.txt -o output_name -w" << std::endl;
return 1;
std::cout << "Prediction mode: full model" << std::endl;
}

Exondata exonin;
Expand Down Expand Up @@ -133,7 +163,7 @@ int main(int argc, char *argv[])
DH_pre_idx3[j] = new int[DH_num3]();
}

//read in modle file
//read in modle file
if (ReadinModel(libfile, quantile_in, exon_mean, exon_sd, coef, DNase_mean, DNase_sd, pre_idx, TC_id, cluster_idx, select_loci, predictor_size, var_size, loci_size, dis_matrix, DH_cluster, DH_coef1, DH_coef2, DH_coef3, DH_pre_idx1, DH_pre_idx2, DH_pre_idx3, DH_num1, DH_num2, DH_num3))
{
return 1;
Expand All @@ -144,7 +174,7 @@ int main(int argc, char *argv[])
return 1;
}
std::cout << "Processing data..." << std::endl;

//check gene expression data
if (CheckTCid(TC_id, exonin.TC_name,predictor_size)) //check if the input file contains the same predictor as library.
{
Expand Down Expand Up @@ -179,29 +209,39 @@ int main(int argc, char *argv[])
data_norm[i][j] = log2(exonin.data[j][i] + 1);
}
QuantileNorm(data_norm[i], quantile_in, predictor_size); //quantile normalization accroding to quantile from UW exon data.

}

StandardizeRow(data_norm, exon_mean, exon_sd, predictor_size, sample_size); //standardize gene expression data
ClusterMean(data_norm, data_mean, cluster_idx, predictor_size, cluster_size, sample_size); //get gene expression cluster mean;

//Locus level prediction
Regression(data_mean, output, coef, pre_idx, var_size, loci_size, sample_size);

//DH cluster level prediction
Regression(data_mean, DH_pre1, DH_coef1, DH_pre_idx1, var_size, DH_num1, sample_size);
Regression(data_mean, DH_pre2, DH_coef2, DH_pre_idx2, var_size, DH_num2, sample_size);
Regression(data_mean, DH_pre3, DH_coef3, DH_pre_idx3, var_size, DH_num3, sample_size);

//Combine result from locus level and DH cluster level
ModelAverage(output, DH_pre1, DH_pre2, DH_pre3, dis_matrix, DH_cluster, loci_size, sample_size);

//convert predicted value back to original scale

if(locus_model != 1)
{
//Locus level prediction
Regression(data_mean, output, coef, pre_idx, var_size, loci_size, sample_size);

//DH cluster level prediction
Regression(data_mean, DH_pre1, DH_coef1, DH_pre_idx1, var_size, DH_num1, sample_size);
Regression(data_mean, DH_pre2, DH_coef2, DH_pre_idx2, var_size, DH_num2, sample_size);
Regression(data_mean, DH_pre3, DH_coef3, DH_pre_idx3, var_size, DH_num3, sample_size);

//Combine result from locus level and DH cluster level
ModelAverage(output, DH_pre1, DH_pre2, DH_pre3, dis_matrix, DH_cluster, loci_size, sample_size);
std::cout << "Using full model for prediction." << std::endl;
}
else
{
//Locus level prediction
Regression(data_mean, output, coef, pre_idx, var_size, loci_size, sample_size);
std::cout << "Using locus-level model for prediction." << std::endl;
}

//convert predicted value back to original scale
StandardizeRow_r(output, DNase_mean, DNase_sd, loci_size, sample_size);

//write output file
//write output file
std::cout << "Writing output file..." << std::endl;
if (WriteWIG(output, select_loci, exonin.sample_name, outfile, bin_size, loci_size, sample_size, write_flag))
if (WriteWIG(output, select_loci, exonin.sample_name, outfile, bin_size, loci_size, sample_size, write_flag, up_bound))
{
return 1;
}
Expand Down

0 comments on commit b039ab3

Please sign in to comment.