scraps

scraps extracts mRNA polyadenylation sites from "TVN"-primed single-cell RNA-seq libraries at near-nucleotide resolution.

scraps (Single Cell RNA PolyA Site Discovery) is currently implemented as a Snakemake pipeline for 10X Genomics 3' end v2/3 libraries (and other platforms with similar library structure, including Drop-seq, Microwell-seq, and BD Rhapsody). If long Read1 is available (estimated ~6% of SRA-deposited data, or now planning new experiments), positional information will be calculated from paired realignment; otherwise, the less optimal anchored Read2 approach is used. scraps will eventually be expanded for analyzing a range of RNA processing changes in single-cell RNA-seq data.

For additional discussions and usage cases, please see bioRxiv preprint.

---

• Example usage
• Supported scRNA-seq platforms
• Output
• Setup
• Dependencies
• Extended function

Example usage

scraps requires the following as input (defined in config.yaml and sample_fastqs.tsv):

• 10X Genomics 3' v2/3 single-cell FASTQs or other platforms (with names "_R1.fastq.gz"" and "_R2.fastq.gz"")
• A STAR genome index (must be generated with STAR 2.7.4a and above)
• Whitelist for cell barcodes (optional but recommended to speed up run time)
• A featureCounts reference (SAF-formatted polya_db, hg38 and mm10 files are included in ref subdirectory)

To run test data, simply execute:

snakemake --snakefile Snakefile \
  --configfile config.yaml \
  --resources total_impact=5 \
  --keep-going

DAG steps illustration

submit jobs in cluster mode

Notes: total_impact is set to 5 for each sample, change this to control how many samples are processed in parallel

---

Supported scRNA-seq platforms

Platform	Library (BC+UMI+A)	Setting	Test data
10x Chromium V3	16 + 12 + 30	chromiumV3	✓
10x V3 - Ultima Genomics	adapter + 16 + 9 + 3 ignored + 8	chromiumV3UG
10x Chromium V2	16 + 10 + 30	chromiumV2	✓
10x Chromium Visium	16 + 10 + 30	visium
Drop-seq	12 + 8 + 30	dropseq	✓
Microwell-seq	6x3 + 6 + 30	microwellseq	✓
BD Rhapsody	9x3 + 8 + 18	bd
inDrop	8 + 6 + 18	indrop

Custom chemistry supported, by editing chemistry.json. Also see synthetic FASTQ tool.

---

Output

1. bedgraph : TVN-priming site pileup

chr11   215106  215107  1
chr11   689216  689217  1
chr11   812862  812863  1
chr11   812870  812871  2
chr11   812871  812872  2

2. count table : +-10 around PolyA_DB sites, by cell barcode

gene    cell    count
AC135178.2_NA_ENSG00000263809_chr17_8377523_-_Intron,RPL26_6154_ENSG00000161970_chr17_8377523_-_3'UTR(M)        AACTCCCGTTCCTCCA        1
AC135178.2_NA_ENSG00000263809_chr17_8377523_-_Intron,RPL26_6154_ENSG00000161970_chr17_8377523_-_3'UTR(M)        CCCATACGTTAAAGAC        1
AC135178.2_NA_ENSG00000263809_chr17_8377523_-_Intron,RPL26_6154_ENSG00000161970_chr17_8377523_-_3'UTR(M)        CGTCCATTCGACAGCC        1
ACTG1_71_ENSG00000184009_chr17_81509999_-_3'UTR(M)      ACATCAGGTGATGTCT        1
ADRM1_11047_ENSG00000130706_chr20_62308862_+_3'UTR(M)   CAGCGACTCTGCCCTA        1

3. html report : various metrics from steps in the pipeline

R functions available for importing results into Seurat object, and finding differential PA site usage. Alternatively, a package of the same functions can be installed with remotes::install_github("rnabioco/scrapR")

---

Setup

1. Clone repository: git clone https://github.com/rnabioco/scraps
2. Check dependencies (ideally with Conda, see below)
3. Place appropriate STAR index in index/ folder, and barcode whitelists in whitelist/
   Download links(all files need to be extracted): GRCh38 index; 10x V2 barcodes; 10x V3 barcodes
4. Edit settings in config.yaml
5. List files in sample_fastqs.tsv, note that SRA accessions in the form of SRR9887775 are supported for direct download
6. Run! (sample results can be found at inst/test_output/)

---

Dependencies

scraps requires the following executables in your PATH:

• Python 3 (developed with version 3.8.5)
• Snakemake (developed with version 3.11.2)
• UMI-tools (developed with version 1.1.1)
• cutadapt (developed with version 3.4)
• STAR (developed with version 2.7.9a)
• Samtools (developed with version 1.3.1)
• Bedtools (developed with version 2.30.0)
• Subread (developed with version 1.6.2)
• MultiQC (developed with version 1.9)

Alternatively, we recommend using Conda to manage these dependencies, simply with: conda env create -f scraps_conda.yml and then conda activate scraps_conda

Docker image for automated deployment can also be found at https://hub.docker.com/r/rnabioco/scraps.

Please also see the Snakemake documentation for general information on executing and manipulating snakemake pipelines.

---

Extended function

1) Measuring internal priming as indicator of apoptotic cytoplasmic poly(A) RNA decay

(Based on widespread RNA decay during apoptosis: Liu and Fu et al.) Use SAF (hg38 version provided in ref subdirectory) file marking all gene regions (5'UTR, intron, CDS, 3'UTR), and helper R functions to process output. Please see Rmarkdown notebook for more.

---

2) Accurate intron/exon quantification for RNA velocity

(See discussions on quantification approaches and pitfalls: Soneson et al.)

Consideration	scraps
Avoid feature double-counting	✓
Take strandedness into account	✓
Avoid count substraction	✓
Resolve spliced vs unspliced target	✓
Speed	✓