Library

get_geno_data(file, datatype)

returns the genotype data. Will skip every other column because genotype probability is duplicated.

get_pheno_data(file, datatype; transposed)

returns the phenotype data. If transposed=true, then the data will be transposed.

scan(Y, G)
scan(Y, G, X; maf_threshold, export_matrix, usegpu, lod_or_pval, timing_file)

This function is the main API for eQTL scans. EQTL scan process includes If no covariate: - filter maf if the threshold is greater than 0 - standardizing phenotype matrix (Y) and genotype matrix (G) - calculate correlation (R) matrix. - computes log of odds (LOD) score matrix, or p-value if lod_or_pval is set to lod - calculate maximum LOD score if export_matrix is set to false.

If covariates exists:
- calculate px 
- computes Y hat and G hat by matrix multiplication. 
- substract Y hat and G hat from Y and G respectively. 
- filter maf if the threshold is greater than 0
- standardizing phenotype matrix (Y) and genotype matrix (G)
- calculate correlation (R) matrix. 
- computes log of odds (LOD) score matrix, or p-value if `lod_or_pval` is set to `lod`
- calculate maximum LOD score if `export_matrix` is set to false.

Arguments:

• Y : a matrix of phenotypes.
• G : a matrix of genotypes.
• X : a matrix of covariates. Default is nothing. If nothing, scan is run without covariates.
• maf_threshold: a floating point number to indicate the maf_threshold. Default is 0.05. Set to 0 if no maf filtering should be done.
• export_matrix : a boolean value that determines whether the result should be the maximum value of LOD score of each phenotype and its corresponding index, or the whole LOD score matrix.
• usegpu : a boolean value that indicates whether to run scan function on GPU or CPU. Default is false, which runs scan on CPU.
• lod_or_pval: a string value of either lod or pval to indicate the desired output.
• timing_file: a string that indicates the file location for the timing outputs. Default is nothing.

Output:

returns LOD score or pval, in vector or matrix format depending on value of export_matrix.

calculate_nr(a, b)

Computes correlation matrix (R) * n. n is not removed because of performance choice.

Arguments

• a : standardized phenotype matrix
• b : standardized genotype matrix

Output:

returns correlation matrix R

calculate_nr(a, b)

Computes correlation matrix (R) * n on GPU with cuBLAS library. n is not removed because of performance choice.

Arguments

• a : standardized phenotype matrix
• b : standardized genotype matrix

Output:

returns correlation matrix R on GPU device (of type CuArray).

calculate_px(x)

Arguments

• x : input matrix

Output:

returns px

cpurun(pheno, geno)
cpurun(pheno, geno, X; maf_threshold, export_matrix, lod_or_pval, timing_file)

Arguments:

• Y : a matrix of phenotypes.
• G : a matrix of genotypes.
• X : a matrix of covariates. Default is nothing. If nothing, scan is run without covariates.
• maf_threshold: a floating point number to indicate the maf_threshold. Default is 0.05. Set to 0 if no maf filtering should be done.
• export_matrix : a boolean value that determines whether the result should be the maximum value of LOD score of each phenotype and its corresponding index, or the whole LOD score matrix.
• lod_or_pval: a string value of either lod or pval to indicate the desired output.
• timing_file: a string that indicates the file location for the timing outputs. Default is nothing.

Output:

returns LOD score or pval, in vector or matrix format depending on value of export_matrix.

filter_maf(genotype; maf_threshold)

Filter genotype data with a minor allele frequency threshold.

Arguments

• genotype : genotype matrix
• maf_threshold : default value is 0.05.

Output:

returns filtered genotype matrix.

find_max_idx_value(lod)

Computes the index of maximum, and maximum value of each row of a matrix. Optimized with multi-threading.

!Notes: Set the thread number with env JULIANUMTHREADS to your desired number of threads. For example: JULIA_NUM_THREADS=16 julia

Arguments

• lod : input matrix.

Output:

returns a matrix with two columns, first column is the index of maximum, second column is the maximum value.

get_pheno_block_size(n, m, p, datatype)

Computes the number of blocks for chuncking large phenotype matrix, to fit into GPU memory.

Arguments

• n : number of individuals
• m : number of phenotypes
• p : number of genotype markers
• datatype: datatype of phenotype. (Float32, or Float64)

Output:

returns

• num_block : number of blocks to disect phenotype
• block_size : size of each block.

get_standardized_matrix(mat)

Computes standardized matrix on CPU utilizing multi-threads.

Arguments

• m : a matrix to be standardized

Output:

returns the standardized matrix of m

get_standardized_matrix_gpu(m)

Computes standardized matrix on GPU. Theoretically this function works with CPU as well. But it is not as fast as the get_standardized_matrix function, which utilizes multi threads on CPU.

Arguments

• m : a matrix to be standardized

Output:

returns the standardized matrix of m

gpu_square_lod(d_nr, n, m, p, export_matrix)

Sets up GPU environment and calls GPU kernels.

Arguments

• d_nr : Output matrix (correlation matrix) in CuArray format
• n : number of individuals
• m : number of phenotypes
• p : number of genotype markers
• export_matrix : boolean value to indicate user desire output

Output:

returns correlation matrix R, or maximum of R if export_matrix is set to false

gpurun(pheno, geno)
gpurun(pheno, geno, X; maf_threshold, export_matrix, timing_file)

Arguments:

• Y : a matrix of phenotypes.
• G : a matrix of genotypes.
• X : a matrix of covariates. Default is nothing. If nothing, scan is run without covariates.
• maf_threshold: a floating point number to indicate the maf_threshold. Default is 0.05. Set to 0 if no maf filtering should be done.
• export_matrix : a boolean value that determines whether the result should be the maximum value of LOD score of each phenotype and its corresponding index, or the whole LOD score matrix.
• lod_or_pval: a string value of either lod or pval to indicate the desired output.
• timing_file: a string that indicates the file location for the timing outputs. Default is nothing.

Output:

returns LOD score, in vector or matrix format depending on value of export_matrix. P value for GPU is not currently implementated.

is_corr_in_range(r, min, max)

Checks whether correlation matrix is in range. Only runs if DEBUG flag is turned on from julia commandline.

Arguments

• r : a matrix to be standardized
• min : minimum value of range
• max : maximum value of range

Output:

errors message will show if correlation matrix is not in range.

lod2p(lod)

Computes p-value based on log of odds score.

Arguments

• lod : LOD matrix

Output:

returns p-value.

lod_kernel(nr, MAX, n)

GPU kernel that computes LOD scores.

Arguments

• nr : correlation matrix. will be updated with output results. (Type CuArray)
• MAX : number of total GPU threads, also called ndrange.
• n : number of individuals

Output:

Results written back to input matrix nr. Returns the LOD scores.

lod_score_multithread(m, nr)

Computes log of odds (LOD) score. Optimized for correlation matrix type is Float32 (single precision).

!Notes: Set the thread number with env JULIANUMTHREADS to your desired number of threads. For example: JULIA_NUM_THREADS=16 julia

Arguments

• m : number of individuals.
• nr : correlation matrix R times n. N will be removed during this step.

Output:

returns LOD score.

lod_score_multithread(m, nr)

Computes log of odds (LOD) score. Optimized for correlation matrix type is Float64 (double precision).

!Notes: Set the thread number with env JULIANUMTHREADS to your desired number of threads. For example: JULIA_NUM_THREADS=16 julia

Arguments

• m : number of individuals.
• nr : correlation matrix R times n. N will be removed during this step.

Output:

returns LOD score.

reduce_kernel(input, rows, cols)

Computes the maximum LOD score of each row. Results written back to the first two columns.

Arguments

• input : LOD score
• rows : number of rows for input
• cols : number of cols for input

Output:

Output is written to the input matrix, first column of input matrix is the index of maximum, and second column is the maximum value of that row.

set_blas_threads(nthread)

Sets the number of threads that will be used by BLAS library. !Notes: Set the thread number with env JULIANUMTHREADS to your desired number of threads. For example: JULIA_NUM_THREADS=16 julia

Arguments

• nthread : desired number of threads

Output:

Shows number of threads set.