Fullscope-seq

A C++ toolset for long-read spatial transcriptomic data (Stereo-seq with ONT/Pacbio/Cyclone): from raw FASTQ to CID mapping in one command.

---

Overview

Fullscope-seq is a high-performance pipeline that turns Stereo-seq long-read spatial transcriptomic FASTQ into cell-resolution CID maps.
Everything is written in C++17 and engineered for Linux servers with ≥32 GB RAM.

---

Quick Start (5 commands)

1. Clone

   git clone https://github.com/wwei-lab/Fullscope.git
   cd Fullscope

2. Run the complete pipeline

   ./Fullscope-1.0 count \
     input.fq adapters.fa anchor.fa 0.8 \
     genes.gtf genome.fa index.cid index_threshold.txt \
     7 6 16 ./output sample1

3. Done – results are in output/

   output/
   ├── Segment/           # demultiplexed reads
   ├── Alignment/         # sorted & indexed BAM
   ├── CIDextract/        # per-read CID tables
   ├── CIDmap/            # CID → spatial index
   └── sample1.summary.txt
   

---

System Requirements

Component	Requirement
OS	Linux (kernel ≥3.10)
Compiler	g++ ≥7 / clang++ ≥5 (C++17)
Third-party	minimap2, samtools in $PATH
Hardware	≥32 GB RAM, ≥16 cores recommended

---

Installation

# 1. clone
git clone https://github.com/wwei-lab/Fullscope.git && cd Fullscope

# 2. build (CMake ≥3.14)
mkdir build && cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make -j$(nproc)
# binary is now ./Fullscope-1.0

---

Step-by-step (if you prefer modular runs)

Step	Command
1. FASTQ segmentation	./Fullscope-1.0 process_fq input.fq adapters.fa anchor.fa 0.8 16 segmented.fq
2. Alignment	minimap2 -K500m --secondary=no -a -x splice --splice-flank=yes -t 16 genome.fa segmented.fq | samtools sort -o aligned.bam
3. CID extraction	./Fullscope-1.0 extract aligned.bam genes.gtf extracted_cid.tsv 16
4. CID mapping	./Fullscope-1.0 map extracted_cid.tsv index.cid index_threshold.txt mapped_cid.tsv 16 7 6

---

Parameter Bible

count – single-command pipeline

./Fullscope-1.0 count \
  <input.fq> <adapters.fa> <anchor.fa> <seg_thresh> \
  <ref.gtf> <ref.fa> <index.cid> <index_thresh.txt> \
  <k> <buckets> <threads> <outDir> <prefix>

Param	Meaning	Example
seg_thresh	segment confidence (0–1)	0.8
k	k-mer length	7
buckets	hash buckets	6
threads	CPU cores	16

---

Auxiliary tools

Tool	One-liner
process_fq	./Fullscope-1.0 process_fq in.fq adapters.fa anchor.fa 0.8 16 out.fq
extract	./Fullscope-1.0 extract aligned.bam genes.gtf cid.tsv 16
extract_fq	./Fullscope-1.0 extract_fq in.fq cid.tsv 16
map	./Fullscope-1.0 map cid.tsv index.cid thresh.txt out.tsv 16 7 6
map_p (precise)	./Fullscope-1.0 map_p cid.tsv index.cid out.tsv 16 7
build_idx	fast: ./Fullscope-1.0 build_idx f list.txt 16 7 6 index.cid
precise: ./Fullscope-1.0 build_idx p list.txt 16 7 index_precise.cid
bamtoref	./Fullscope-1.0 bamtoref genes.gtf in.bam ref_list.txt out 16 T

---

Output Structure

output/
├── Segment/           # demultiplexed *.fq
├── Alignment/         # *.bam + *.bai
├── CIDextract/        # *.cid.tsv
├── CIDmap/            # *.mapped.cid.tsv
└── <prefix>.summary.txt

---

Notes & Pro-tips

1. Dependencies – ensure minimap2 and samtools are in $PATH.
2. Memory – 1 GB per 1 M reads is a safe rule of thumb.
3. Threads – leave 2 cores free for I/O.
4. Disk – reserve 5× the input FASTQ size for intermediates.

---

Help & Version

./Fullscope-1.0