TIF-Seq2

The analysis pipeline and downstream analysis in jingwen/TIFseq2 are used for publication TIF-Seq2 disentangles overlapping isoforms in complex human transcriptomes. The scripts for TIF-Seq2 data pre-processing and alignment in other TIF-Seq2 related publication are in PelechanoLab/TIFseq2.

Prepare sample sheet for demultiplexing

prep_sampleSheet.awk <index_input.txt> > <sample_sheet>

Demultiplex

bcl2fastq -R <input_dir> -o <fastq_dir> --sample-sheet <sample_sheet> --no-lane-splitting --barcode-mismatches demultiplex_stats.awk <fastq_dir>/Stats/DemultiplexingStats.xml > <fastq_dir>/demultiplex_stat.txt

Preprocess

preprocess.sh -I <fastq_dir> -O <output_dir> -j <thread_number> -A <polyA_length>

Align TIF-Seq2 reads and remove PCR duplicates

STAR_align.sh -R <STAR_index_dir> -A <splicing_junction_gtf> -I <output_dir>/cutPolyA -O <STAR_output_dir> -p <thread_number> -j <max_intron_size> -m 0

Fetch boundaries of transcription isoforms (TIFs)

python boundary.py <input_bam>

Filter internal priming of 3'end

Rscript clean_As.R <3end_ctss_path>

Cluster 5' ends and 3' ends respectively

Rscript cluster_end.R <TIF_5end_path> <TIF_3end_path> <3'T-fill_3end_path>

Construct transcription isoform (TIF) boundaries

Rscript form_TIF.R