Contact: Seyla Wickramasinghe (C3G Services Intern 2025) seyla.wickramasinghe@mail.mcgill.ca, seyla.wick@gmail.com

INSTALLATION
Install Alevin-Fry via GitHub: https://github.com/COMBINE-lab/alevin-fry
$ conda install -c bioconda alevin-fry

SETUP
Create a working directory, ex: 
export AF_SAMPLE_DIR=$PWD/af_test_workdir
mkdir $AF_SAMPLE_DIR
cd $AF_SAMPLE_DIR

Download or move fastqs to correct folder
FASTQ_DIR="$AF_SAMPLE_DIR/data/fastq_directory"
mkdir -p $FASTQ_DIR

Download or move reference to correct folder
REF_DIR="$AF_SAMPLE_DIR/data/refdata-gex-GRCh38-2024-A”
mkdir -p $IDX_DIR/ref

RUN
01 Run alevin fry and cellranger
02 Process samples
03 Integrate samples together
04 Elbow plot to decide number of principal components
05 Clustering and umap

ANNOTATION
06 Marker trimming through visualizaton of Feature and ViolinPlots
07 Manual annotation - creating .csv with cluster -> label mapping using AddModuleScore
07 De novo annotation - creating .csv with top 10 genes per cluster
08 Removal of clusters 15-19 (not annotating well after de novo annotatation)
09 Final annotated umap

FURTHER ANALYSIS
10-01 Calculate processing times for alevin fry using log files
10-01 Calculate processing times for cellranger using log files
10-02 Generate correlation plot for runtimes between alevin fry and cellranger
11 Generate correlation and QC ViolinPlots for select metrics
12 Generate stacked barplot for proportion of each label per cluster
13-01 Verify if novel fibroblast subpopulation can still be picked up with alevin fry
13-02 Verify if mature dendritic cell immune subset can still be picked up with alevin fry