TripletLogger v1.0

Alignment-free estimation of CAG / CTG repeat lengths from targeted amplicon sequencing data (Illumina MiSeq short reads or Oxford Nanopore long reads).

Github page- https://github.com/hchetia/TripletLogger

TripletLogger scans FASTQ reads directly with an N-aware regular expression, counts pure target triplets within each accepted match, and reports per-read repeat lengths, a raw repeat-length frequency table, summary metrics, and an estimated allele call (kernel-density peak picking). No alignment to a reference is required.

---

Features

• Supports both CAG and CTG repeats with triplet-specific logic (regex, counting rule, flank trimming, Type B handling).
• Works on short reads (MiSeq) and long reads (ONT) via a single --readType switch with sensible defaults for each.
• Tolerates sequencing errors through configurable Type A (total non-target) and Type B (consecutive non-target) thresholds, with a length-adaptive Type A floor.
• Trims absorbed downstream flank tracts (CCG-like for CAG, CGG-like for CTG).
• Streams FASTQ in chunks — memory-safe for large ONT runs.
• Outputs per-sample CSVs: repeat metrics, reads-per-repeat-length, and rejected-match sizes.

Requirements

• R ≥ 4.0
• Bioconductor: ShortRead, Biostrings
• CRAN: dplyr, optparse, tictoc

Install:

install.packages(c("dplyr", "optparse", "tictoc"))
if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager")
BiocManager::install(c("ShortRead", "Biostrings"))

Usage

Rscript TripletLogger_v1_0.R -i <input.fastq.gz> [options]

Required

Flag	Description
-i, --input	Input FASTQ file (plain or gzipped)

Common options

Flag	Default	Description
-o, --output	.	Output directory
-t, --tripletType	CAG	Repeat type: CAG or CTG
-r, --readType	long	long (ONT) or short (MiSeq)
-q, --qThreshold	20	Minimum mean base quality per read
-s, --sampleId	filename	Override sample ID

Advanced options

Flag	Default	Description
-m, --minRepeatLength	1	Minimum repeat length to report
--maxTypeA	5	Fixed floor for Type A errors (total non-target triplets)
--maxTypeB	1 (CAG) / 2 (CTG)	Max consecutive non-target triplets
--typeArate	0.05	Per-triplet error tolerance for adaptive Type A
--chunkSize	20000	FASTQ streaming chunk size
--alleleBW	1.5	Kernel density bandwidth for allele calling
--freqRangeMin	1	Minimum repeat length for frequency table
--freqRangeMax	1000	Maximum repeat length (short reads capped at 120)

Examples

CAG repeats from a Nanopore run:

Rscript TripletLogger_v1_0.R -i sample01.fastq.gz -o results/ -t CAG -r long

CTG repeats from a MiSeq run:

Rscript TripletLogger_v1_0.R -i sample02.fastq.gz -o results/ -t CTG -r short

Output

For each input FASTQ, TripletLogger writes three CSVs to the output directory:

• <sample>_<TYPE>TRIPLETLogger.v1.0.<readtype>_RepeatMetrics.csv — summary metrics and estimated allele call(s).
• <sample>_<TYPE>TRIPLETLogger.v1.0.<readtype>_NumReadsPerRepeat.csv — raw repeat-length frequency distribution.
• <sample>_<TYPE>TRIPLETLogger.v1.0.<readtype>_ThresholdRejected.csv — sizes of matches rejected by error thresholds.

Version

v1.0

License

MIT