Problems in predict_variants.R script

[dreammxy]

Hi,
I have two questions:

1. As you mentioned, this software used GATK best practise vcf file as input, but in your pipeline it is also used bam file. So this bam file in which step GATK best practise created? or just first step mapping bam is OK?
2. I have run to predict_variants.R script. It always print "Error: both model and input file need to be specified.
   Execution halted". Below is output files before this step. Could you help how to correct it?
   Thank you!
   

[mbosio85]

Hi,

about point 1, any bam file should be enough, and about point 2, can you share with me how do you run the pipeline?
Do you use nextflow implementation or the shell script ?

The error states that either the prediciton model is not defined, or the input file for the R script is not defined.
From lines 99-106 of the shell script :
if $MODEL is not defined at the beginning of the script, you may get the error from above

echo '-----------------------------------------------------'
echo 'Classification' 
echo '-----------------------------------------------------'
echo ' -- Check log : ' $RESFOLDER/"Classification.log"
Rscript $REPOSITORY_PATH/predict_variants.R \
      --model $MODEL \
      --input $RESFOLDER/$OUTPREFIX"_postprocessed.csv"  \
      --out $RESFOLDER/$OUTPREFIX"_classified.tmp.csv"  > $RESFOLDER/"Classification.log" 2>&1