Analysis for and development of the Tetrahymena Gene Network Explorer
Navigate to the top level directory and run one of the following:
Mac – Apple Silicon (Rosetta required):
bash build_tgne_env_silicon.sh
Linux or Mac – Intel:
bash build_tgne_env.sh
Download the [precomputed data](LINK # FIXME).
Create necessary directories within the top level directory:
mkdir -p input_data/eggnog_annotations_file/
mkdir -p input_data/interproscan_tsv_file/
mkdir -p active_files/
Place eggnog_compiled_2024_jun28.annotations in the input_data/eggnog_annotations_file/ directory.
Place interproscan_compiled.tsv in the input_data/interproscan_tsv_file/ directory.
Place allgood_filt_agg_tidy_2021aligned_qc_rma_expression_full.csv in the active_files/ directory.
Place rna_seq.csv in the active_files/ directory.
bash s-execute-pipeline.sh pipeline_precomputed.txt
- Kallisto:
Create necessary directories within the top level directory:
mkdir -p input_data/kallisto_data_folders
Build the conda environment.
bash build_tgne_env.sh
Activate the conda environment:
conda activate tgne.env
Install the SRA Toolkit. Ensure that the toolkit is in your system path.
Download the RNA-seq data from SRA (this will take a few hours):
prefetch PRJNA861835
The declare -A flag requires bash version > 4.
declare -A time_points=(
[SRR20576712]="240min_B"
[SRR20576713]="240min_A"
[SRR20576714]="210min_B"
[SRR20576715]="210min_A"
[SRR20576716]="180min_B"
[SRR20576717]="180min_A"
[SRR20576718]="150min_B"
[SRR20576719]="150min_A"
[SRR20576720]="120min_B"
[SRR20576721]="120min_A"
[SRR20576722]="090min_B"
[SRR20576723]="090min_A"
[SRR20576724]="060min_B"
[SRR20576725]="060min_A"
[SRR20576726]="030min_B"
[SRR20576727]="030min_A"
[SRR20576728]="000min_B"
[SRR20576729]="000min_A"
)
for i in {12..29}; do
srr_id="SRR205767${i}"
time_sample="${time_points[$srr_id]}"
fasterq-dump -e ${num_cores} -o ${time_sample}_# -S $srr_id
gzip ./${time_sample}*.fastq
done
Or (if you have pigz installed):
num_cores=<NUMBER OF CORES TO USE>
declare -A time_points=(
[SRR20576712]="240min_B"
[SRR20576713]="240min_A"
[SRR20576714]="210min_B"
[SRR20576715]="210min_A"
[SRR20576716]="180min_B"
[SRR20576717]="180min_A"
[SRR20576718]="150min_B"
[SRR20576719]="150min_A"
[SRR20576720]="120min_B"
[SRR20576721]="120min_A"
[SRR20576722]="090min_B"
[SRR20576723]="090min_A"
[SRR20576724]="060min_B"
[SRR20576725]="060min_A"
[SRR20576726]="030min_B"
[SRR20576727]="030min_A"
[SRR20576728]="000min_B"
[SRR20576729]="000min_A"
)
for i in {12..29}; do
srr_id="SRR205767${i}"
time_sample="${time_points[$srr_id]}"
fasterq-dump -e ${num_cores} -o ${time_sample}_# -S $srr_id
pigz -p ${num_cores} ./${time_sample}*.fastq
done
Move the samples to a new folder:
mkdir -p raw/
mv *.fastq.gz raw/
Run FastQC on each of the fastq.gz files:
cd raw
mkdir -p qc/
for filename in *.fastq.gz; do fastqc --threads 16 --outdir ./qc $filename; done
Compile the FastQC results with MultiQC:
multiqc ./qc/ -o ./qc/
Download the proper Trimmomatic adapter sequence:
curl -O https://raw.githubusercontent.com/usadellab/Trimmomatic/refs/heads/main/adapters/TruSeq3-PE.fa
Run Trimmomatic on each pair of fastq.gz files:
output_folder="trimmed/"
mkdir -p ${output_folder}
file_name_1=<FILE_NAME_1>
file_name_2=<FILE_NAME_2>
trimmomatic PE ${FASTQ_DIR%/}/${file_name_1} ${FASTQ_DIR%/}/${file_name_2} ${output_folder%/}/p_trimmed_${file_name_1} ${output_folder%/}/up_trimmed_${file_name_1} ${output_folder%/}/p_trimmed_${file_name_2} ${output_folder%/}/up_trimmed_${file_name_2} ILLUMINACLIP:TruSeq3-PE.fa:2:30:10:2:True LEADING:3 TRAILING:3 MINLEN:36 -threads <NUM_THREADS>
Move the unpaired reads to a new folder:
cd trimmed/
mkdir -p up/
mv up_trimmed_* up/
Run FastQC on each of the trimmed fastq.gz files:
mkdir -p qc/
fastqc --threads <NUM_THREADS> --outdir ./qc/ ./<FILE_NAME>
Compile the FastQC results with MultiQC:
multiqc ./qc/ -o ./qc/
Download the latest CDS file:
# Uncomment the two lines below if on a cluster
# mkdir -p cds_index/
# curl -o ./cds_index/cds.fasta https://github.com/yefei521/Tetrahymena_Genome_annotation_V2024/releases/download/V2024.2/Tetrahymena_Genome_annotation_V2024_CDS.fasta
# Comment two three lines below if on a cluster
wget https://github.com/yefei521/Tetrahymena_Genome_annotation_V2024/releases/download/V2024.2/Tetrahymena_Genome_annotation_V2024_CDS.fasta
mv ./Tetrahymena_Genome_annotation_V2024_CDS.fasta ./cds_index/cds.fasta
Index the latest CDS file with Kallisto:
kallisto index --index ./cds_index/kallisto_cds_index ./cds_index/cds.fasta
Run Kallisto on each pair of trimmed fastq.gz files:
output_folder="counts/"
mkdir -p ${output_folder}
file_name_1=<FILE_NAME_1>
file_name_2=<FILE_NAME_2>
file_basename=basename="${file_name_1%_*}"
kallisto quant -t {num_cores} -o ${output_folder%/}/kallisto_quant_${file_basename} -i ./cds_index/kallisto_cds_index ${file_name_1} ${file_name_2}
Place each of the resulting Kallisto output folders in the counts/ directory in the input_data/kallisto_data_folders directory.
- Eggnog:
Create necessary directories within the top level directory:
mkdir -p input_data/eggnog_annotations_file
Build the conda environment.
Activate the conda environment:
conda activate tgne.env
Download the latest protein sequences file:
curl -o ./pep.fasta https://github.com/yefei521/Tetrahymena_Genome_annotation_V2024/releases/download/V2024.2/Tetrahymena_Genome_annotation_V2024_Protein_addAnno.fasta
Prepare the latest protein sequences file:
Place scripts/eggnog_pep_prep.py in your current directory.
python eggnog_pep_prep.py
Download the Eggnog database:
mkdir -p eggnog_data/
download_eggnog_data.py --data_dir eggnog_data/ -P -H -d 2759
Respond to the prompts:
Download main annotation database? [y,n] y
Download taxa database? [y,n] y
Download diamond database (~4GB after decompression)? [y,n] n
Download pfam database (~3GB after decompression)? [y,n] y
Download HMMER database of tax ID 2759? [y,n] y
Please, specify a non-empty name for the database (e.g. Bacteria) [default:2759]: 2759
Use Eggnog to gather ortholog hits:
emapper.py --data_dir /eggnog_data -m hmmer -d 2759 --no_annot -i ./pep.fasta -o pep_a --cpu 12 --tax_scope Eukaryota --target_taxa 2759 --report_orthologs --report_no_hits --go_evidence non-electronic --pfam_realign none --dbtype hmmdb --override
Use Eggnog to gather orthologous annotations:
emapper.py --data_dir /project2/apturkew/michael/eggnog_data -m no_search --annotate_hits_table pep_a.emapper.seed_orthologs -o pep_b --dbmem --excel --override
Place the resulting Eggnog output file, pep_b.emapper.annotations, in the input_data/eggnog_annotations_file/ directory.
- InterProScan:
Create necessary directories within the top level directory:
mkdir -p input_data/eggnog_annotations_file
Install InterProScan (Linux only).
Build the conda environment.
Activate the conda environment:
conda activate tgne.env
Download the latest protein sequences file:
curl -o ./pep.fasta https://github.com/yefei521/Tetrahymena_Genome_annotation_V2024/releases/download/V2024.2/Tetrahymena_Genome_annotation_V2024_Protein_addAnno.fasta
Prepare the latest protein sequences file:
Place scripts/interproscan_pep_prep.py in your current directory.
python interproscan_pep_prep.py
path_to_interproscan=<PATH TO AND INLCUDING interproscan-5.68-100.0/>
${path_to_interproscan%/}/interproscan.sh -i pep_cleaned.fasta -f tsv -d ./ -cpu <NUM_CPUS>
Place the resulting InterProScan output file, pep_cleaned.fasta.tsv, in the input_data/eggnog_annotations_file/ directory.
- Build the TGNE:
Create necessary directories within the top level directory:
mkdir -p active_files/
bash s-execute-pipeline.sh pipeline_precomputed.txt
Build the TGNE following "Building the TGNE from precomputed data" or "Building the TGNE from raw data."
Navigate to the top level directory and run the following:
bash s-execute-pipeline.sh ./regranulation/regranulation_pipeline.txt