MKL-related error during installation

[xiangzhu]

We tried to run the testing example provided in the paper but we came
the following MKL-related error:

2021-07-13 19:57:38,225 | qstats.revreg | DEBUG | Rank 0 using 25 masks on 27 SNPs [0 to 26]
INTEL MKL ERROR:
/opt/aci/sw/anaconda3/2020.07_gcc-4.8.5-bzb/lib/libmkl_avx.so:
undefined symbol: mkl_sparse_optimize_bsr_trsm_i8.
Intel MKL FATAL ERROR: Cannot load libmkl_avx.so or libmkl_def.so.

Regarding the Tejaas code, here is what I ran

[olb5053@comp-sc-0224 ~]$ conda create -n tejaas_testing -y 
[olb5053@comp-sc-0224 ~]$ source activate tejaas_testing
(tejaas_testing) [olb5053@comp-sc-0224 ~]$ module load mkl
(tejaas_testing) [olb5053@comp-sc-0224 ~]$ module load anaconda3
(tejaas_testing) [olb5053@comp-sc-0224 ~]$ module load python
(tejaas_testing) [olb5053@comp-sc-0224 ~]$ conda install -c conda-forge mpi4py statsmodels scikit-learn mpmath pygtrie -y
(tejaas_testing) [olb5053@comp-sc-0224 ~]$ cd work/tejaas-master
(tejaas_testing) [olb5053@comp-sc-0224 tejaas-master]$ tests/run_mwe.sh

And here is the output I get:

2021-07-13 19:57:22,697 | utils.args | INFO | Running TEJAAS v1.0.1
2021-07-13 19:57:22,699 | utils.args | INFO | Method: rr
2021-07-13 19:57:22,699 | utils.args | INFO | Null Model: perm
2021-07-13 19:57:22,699 | utils.args | INFO | Sigma_beta: 0.1
2021-07-13 19:57:22,699 | __main__ | DEBUG | Using 1 cores
2021-07-13 19:57:22,762 | iotools.data | DEBUG | Reading expression levels for trans-eQTL discovery
2021-07-13 19:57:22,848 | iotools.data | DEBUG | Reading expression levels for target gene discovery
2021-07-13 19:57:22,938 | iotools.data | DEBUG | Found 500 genes of 462 samples
2021-07-13 19:57:22,939 | iotools.data | DEBUG | Reading gencode file for gene information
2021-07-13 19:57:36,056 | iotools.data | DEBUG | Selecting common samples of genotype and gene expression
2021-07-13 19:57:36,057 | iotools.data | DEBUG | Before expression selection: 500 genes from 462 samples
2021-07-13 19:57:36,575 | iotools.data | DEBUG | Removed 3 SNPs because of non-single letter polymorphisms
2021-07-13 19:57:36,575 | iotools.data | DEBUG | Removed 0 SNPs because of unknown allele symbol
2021-07-13 19:57:36,575 | iotools.data | DEBUG | Removed 12 SNPs because of ambiguous strands
2021-07-13 19:57:36,576 | iotools.data | DEBUG | Removed 0 SNPs because of unknown RSIDs
2021-07-13 19:57:36,576 | iotools.data | DEBUG | Removed 58 SNPs because of low MAF < 0.01
2021-07-13 19:57:36,576 | iotools.data | DEBUG | Removed 0 SNPs because of low MAF (current)
2021-07-13 19:57:36,576 | iotools.data | DEBUG | 27 SNPs after filtering 
2021-07-13 19:57:36,576 | iotools.data | DEBUG | After expression selection: 480 genes from 462 samples
2021-07-13 19:57:36,576 | iotools.data | DEBUG | Retained 462 samples
2021-07-13 19:57:36,576 | iotools.data | DEBUG | Generate cis-masks for GX matrix for each SNP
2021-07-13 19:57:36,581 | iotools.data | DEBUG | Applying KNN correction on gene expression and genotype
2021-07-13 19:57:38,222 | __main__ | DEBUG | After prefilter: 27 SNPs and 480 genes in 462 samples
2021-07-13 19:57:38,224 | __main__ | DEBUG | Sigma_beta fixed to 0.1
2021-07-13 19:57:38,225 | qstats.revreg | DEBUG | Rank 0 using 25 masks on 27 SNPs [0 to 26]

INTEL MKL ERROR:
/opt/aci/sw/anaconda3/2020.07_gcc-4.8.5-bzb/lib/libmkl_avx.so:
undefined symbol: mkl_sparse_optimize_bsr_trsm_i8.
Intel MKL FATAL ERROR: Cannot load libmkl_avx.so or libmkl_def.so.
--------------------------------------------------------------------------
Primary job  terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------

--------------------------------------------------------------------------
mpirun detected that one or more processes exited with non-zero
status, thus causing the job to be terminated. The first process to do so was:

  Process name: [[63996,1],0]

  Exit code:    2

[banskt]

The core of Tejaas is written in C which is compiled using linear algebra routines like MKL or OpenBLAS. The error occurs because the MKL libraries are not found during runtime.

We repackaged our source code in PyPI. Could you try installing it using the detailed installation instruction?