From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> On Behalf Of Taylor, Lisa
Sent: Tuesday, June 13, 2023 2:58 PM
To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] Error in longitudinal process step 3

#!/bin/sh

# NOTE, basic freesurfer recon-all should already be run on subjects before this longitudinal processing step

# see freesurfer_prep.sh script for details

wd=$(pwd)

#dirs=`ls -d1 ${wd}/* | awk -F/ '/'_tp'?[0-9]/ { print $NF } '`

subids=`ls -d1 ${wd}/* | awk -F/ '/'_tp'?[0-9]$/ { print $NF } ' | sed s/'_tp'[0-9]//g | uniq`

echo $subids

if [[ $# -lt 2 ]] ; then

  echo 'Usage: sh freesurfer_longitudinal_step2.sh [queue] [account_to_charge]'

  echo '[queue] = free | standard'

  echo '[account_to_charge] = DBKEATOR | DBKEATOR_LAB | UMAR_LAB'

  echo 'PLEASE ONLY USE THE FREE QUEUE UNLESS OTHERWISE INSTRUCTED'

  exit 0

fi

for subid in ${subids}; do

      #get list of timepoints from directory names

      timepoints=`ls -d1 ${subid}* | cut -d'_' -f2`

      len=`echo "${timepoints}" | wc -l`

      if [ ${len} -gt 1 ]

      then

            printf '#!/bin/bash\n' > freesurfer_${subid}_long_s2.sub

            printf '#SBATCH --job-name=fs_%s_tp%s_long\n' ${subjid} ${time} >> freesurfer_${subid}_long_s2.sub

            printf '#SBATCH -A %s\n' ${2} >> freesurfer_${subid}_long_s2.sub

            printf '#SBATCH -p %s\n' ${1} >> freesurfer_${subid}_long_s2.sub

            printf '#SBATCH --nodes=1\n' >> freesurfer_${subid}_long_s2.sub

            printf '#SBATCH --ntasks=1\n' >> freesurfer_${subid}_long_s2.sub

            printf '#SBATCH --cpus-per-task=1\n' >> freesurfer_${subid}_long_s2.sub

            printf '#SBATCH --error=slurm-%%J.err\n' >> freesurfer_${subid}_long_s2.sub

            printf 'module load freesurfer/6.0.1\n' >> freesurfer_${subid}_long_s2.sub

            printf 'source ${FREESURFER_HOME}/SetUpFreeSurfer.sh\n' >> freesurfer_${subid}_long_s2.sub

            printf 'export SUBJECTS_DIR=%s\n' ${wd} >> freesurfer_${subid}_long_s2.sub

            printf '#run freesurfer longitudinal workflow step 2\n' >> freesurfer_${subid}_long_s2.sub

            printf 'srun ${FREESURFER_HOME}/bin/recon-all -base %s' ${subid} >> freesurfer_${subid}_long_s2.sub

            for time in ${timepoints}; do

                  printf " -tp %s_%s" ${subid} ${time} >> freesurfer_${subid}_long_s2.sub

            done

            printf " -all\n"  >> freesurfer_${subid}_long_s2.sub

      fi

     sbatch freesurfer_${subid}_long_s2.sub

done

Hi Lisa,

I’m sorry that I didn’t make it clear.

From what I understand, your longitudinal processing should look like the following:

In step 1 (cross), you should have processed each of your time points separately, in your case, <tpNid> = 6008_tp1, 6008_tp2, 6008_tp3

recon-all -all -s <tpNid> -i path_to_tpN_dcm

In step 2 (base), recon-all -base <templateid> -tp <tp1id> -tp <tp2id> ... -all

You used 6008 for <templateid>, so your recon-all command should look like ‘recon-all -base 6008 -tp 6008_tp1 -tp 6008_tp2 -tp 6008_tp3 -all’.

I’m wondering if ‘-tp 6008_6008’ was passed to your ‘recon-all -base’ command. I was hoping to take a look of recon-all.log in your base directory - /dfs8/pub/ltaylor2/downs/6008/scripts/recon-all.log.

In step 3 (long), ‘recon-all -long <tpNid> <templateid> -all’. It should output in directory <tpNid>.long.<templateid>.

Best,

Yujing

        External Email - Use Caution