SMultiXcan error - No objects to concatenate

[Zhenyao Ye]

Dear Predixcan team! 

I'm having an issue with SMultiXcan using the elastic net models. 

The error I got is "INFO - Unexpected error: No objects to concatenate". 

./SMulTiXcan.py --models_folder PrediXcan/elastic_net_models/ --models_name_filter "en_Brain_(.*).db" --models_name_pattern "en_Brain_(.*).db" --snp_covariance SMulTiXcan/gtex_v8_expression_elastic_net_snp_smultixcan_covariance.txt.gz --metaxcan_folder SMulTiXcan/metaxcan_spredixcan_folder/ --metaxcan_filter "en_Brain_(.*)_allchr_cpd.csv" --metaxcan_file_name_parse_pattern "en_(.*).csv" --gwas_folder GWASonCPD/en_gwas/ --gwas_file_pattern *.txt --snp_column SNP --effect_allele_column A1 --non_effect_allele_column A2 --beta_column BETA --pvalue_column P --output SMulTiXcan/en_output_chr/SMultixcan_allchr_cpd.csv

INFO - Creating context

INFO - Creating MetaXcan results manager

INFO - Loading genes

INFO - Context for snp covariance

INFO - Assessing GWAS-Models SNP intersection

INFO - Processing GWAS command line parameters

INFO - Unexpected error: No objects to concatenate

I used spredixcan to generate metaxcan files. For example, 

./SPrediXcan.py --model_db_path PrediXcan/elastic_net_models/en_Brain_Substantia_nigra.db --covariance PrediXcan/elastic_net_models/en_Brain_Substantia_nigra.txt.gz --gwas_folder GWASonCPD/en_gwas/ --gwas_file_pattern ".*txt" --snp_column SNP --effect_allele_column A1 --non_effect_allele_column A2 --beta_column BETA --pvalue_column P --output_file SMulTiXcan/metaxcan_spredixcan_folder/en_Brain_Substantia_nigra_allchr_cpd.csv. Is it correct? If not, what is metaxcan files and how to generate them? Because I didn't see the Metaxcan.py script in the "software" directory, which was downloaded from the GutHub MetaXcan. 

Any suggestion is appreciated. Thank you very much for your help! 

Best regards,

Zhenyao

[Alvaro Barbeira]

Hi Zhenyao,

I'm sorry a long time has gone by without any answer to your issue. Hopefully you have solved the problem by now.

For completeness´ sake, I add some comments below.

Best,

Alvaro

On Thu, Jun 24, 2021 at 11:18 AM Zhenyao Ye <yezh...@terpmail.umd.edu> wrote:

Dear Predixcan team! 

I'm having an issue with SMultiXcan using the elastic net models. 

The error I got is "INFO - Unexpected error: No objects to concatenate". 

./SMulTiXcan.py --models_folder PrediXcan/elastic_net_models/ --models_name_filter "en_Brain_(.*).db" --models_name_pattern "en_Brain_(.*).db" --snp_covariance SMulTiXcan/gtex_v8_expression_elastic_net_snp_smultixcan_covariance.txt.gz --metaxcan_folder SMulTiXcan/metaxcan_spredixcan_folder/ --metaxcan_filter "en_Brain_(.*)_allchr_cpd.csv" --metaxcan_file_name_parse_pattern "en_(.*).csv" --gwas_folder GWASonCPD/en_gwas/ --gwas_file_pattern *.txt --snp_column SNP --effect_allele_column A1 --non_effect_allele_column A2 --beta_column BETA --pvalue_column P --output SMulTiXcan/en_output_chr/SMultixcan_allchr_cpd.csv

INFO - Creating context

INFO - Creating MetaXcan results manager

INFO - Loading genes

INFO - Context for snp covariance

INFO - Assessing GWAS-Models SNP intersection

INFO - Processing GWAS command line parameters

INFO - Unexpected error: No objects to concatenate

This is odd. But it looks like a problem in the --metaxcan_file_name_parse_pattern argument.

SMultiXcan doesn´t assume anything about naming conventions, so it must be informed on how to relate S-PrediXcan result files to their originating models the "name parse pattern" arguments define a pattern to use when identifying tissues (and/or phenotype names).

From the command below, I recommend trying with:

--metaxcan_file_name_parse_pattern "en_Brain_(.*) _allchr_cpd.csv"

(you can take a look at the wiki (https://github.com/hakyimlab/MetaXcan/wiki/Command-Line-Reference#smultixcanpy) for an example)

I used spredixcan to generate metaxcan files. For example, 

./SPrediXcan.py --model_db_path PrediXcan/elastic_net_models/en_Brain_Substantia_nigra.db --covariance PrediXcan/elastic_net_models/en_Brain_Substantia_nigra.txt.gz --gwas_folder GWASonCPD/en_gwas/ --gwas_file_pattern ".*txt" --snp_column SNP --effect_allele_column A1 --non_effect_allele_column A2 --beta_column BETA --pvalue_column P --output_file SMulTiXcan/metaxcan_spredixcan_folder/en_Brain_Substantia_nigra_allchr_cpd.csv. Is it correct? If not, what is metaxcan files and how to generate them? Because I didn't see the Metaxcan.py script in the "software" directory, which was downloaded from the GutHub MetaXcan. 

MetaXcan.py  was an old implementation, containing the S-PrediXcan method and other ancillary operations.

S-PrediXcan.py is a clean up of the above focused on the S-PrediXcan gene-level association method. Everything else in MetaXcan.py got either deprecated or adopted in other scripts of MetaXcan repository.

Any suggestion is appreciated. Thank you very much for your help! 

Best regards,

Zhenyao

--
You received this message because you are subscribed to the Google Groups "PrediXcan/MetaXcan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to predixcanmetax...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/predixcanmetaxcan/db2567c6-ec1b-4733-b65e-060393868b9cn%40googlegroups.com.