toolkit_seed.md

root

scarecrow seed

This tool identifies barcode seed positions and potential linker sequences within reads. It returns a CSV file of barcode alignments, a TSV file of nucleotide frequencies for each fastq file, and a TSV file of conserved sequence runs. This step should be repeated for each barcode whitelist. The barcode whitelist is a text file with one barcode sequence per line and no header. The --num_reads option specifies the number of reads to run the analysis on, and defaults to 100000. Using all reads (--num reads 0) is neither necessary nor recommended, the tool does not run multiprocessing so will take considerable time to process millions of reads. Subsets of reads are sampled randomly, the random seed can be set by --random_seed <int> and defaults to 1234.

scarecrow seed --fastqs R1.fastq.gz R2.fastq.gz \
    --num_reads 100000 \
    --out BC1_counts.csv \
    --barcodes BC1:v1:v1_whitelist.txt

The --out file has the below format. The file_index is the 0-based index of the FASTQ file (of files passed via --fastq) on which an exact match was identified; the associated FASTQ file for the index is listed under file. The read_name column indicates the read name of the sequencing read, and seqlen indicates the length of the read. The barcode_whitelist contains the first two elements of the string passed to --barcdoes, barcode is the sequence of the matched barcode, orientation is the orientation in which the sequence was found on the read, start and end are the positions of the alignment, mismatches is the number of mismatching bases between the fastq sequence and barcode.

file_index      file    read_name       seqlen  barcode_whitelist       barcode orientation     start   end     mismatches
1       SRR28867558_2.fastq.gz  SRR28867558.5   74      BC1:n99_v5      CACTTTCA        reverse 8       15      0
1       SRR28867558_2.fastq.gz  SRR28867558.5   74      BC1:n99_v5      GTGCTTGA        reverse 24      31      0
1       SRR28867558_2.fastq.gz  SRR28867558.5   74      BC1:n99_v5      GTGCTTGA        reverse 50      57      0
2       SRR28867558_3.fastq.gz  SRR28867558.18  86      BC1:n99_v5      GTGCTTGA        forward 63      70      0
2       SRR28867558_3.fastq.gz  SRR28867558.18  86      BC1:n99_v5      TGTGTATG        forward 79      86      0

The --out[-csv]_conserved.tsv file has the below format. The file_index column indicates the 0-based FASTQ file index from which the conserved sequence was identified, start and end indicate the conserved sequence positions, length indicates its length, sequence is the conserved sequence, median_frequency is the median frequency of the sequence across the reads. Conserved sequences represent linkers, adapters, and other 'fixed' sequences within a read.

file_index      start   end     length  sequence        median_frequency
2       31      48      18      TCGCATCGGCGTACGACT      0.8837
2       57      78      22      ATCCACGTGCTTGAGACTGTGG  0.8698

The --out[-csv]_read[1|2]_frequencies.tsv file has the below format. The read column indicates the sequencing read analysed, the position indicates the position along the read, A, C, G, T, N indicate the frequencies of the observed nucleotides at that position across reads.

read    position        A       C       G       T       N
read1   1       0.4928  0.1722  0.2761  0.0590  0.0000
read1   2       0.4973  0.1531  0.1740  0.1747  0.0009
read1   3       0.2076  0.1803  0.4372  0.1749  0.0000