SparvusGenome

Scripts and notes for assembling, QC-ing, submitting, and annotating the Staurois parvus genome.

This repository documents the workflow used to generate a new PacBio HiFi genome assembly for the Bornean rock frog / foot-flagging frog, Staurois parvus, screen it for NCBI submission, and annotate it with NCBI EGAPx.

Project overview

The main goals of this workflow are:

1. Generate a new S. parvus genome assembly from PacBio HiFi reads.
2. Assess assembly quality using summary statistics, BUSCO, GC/coverage checks, and contamination screening.
3. Clean the assembly for NCBI submission.
4. Annotate the cleaned genome using NCBI EGAPx with S. parvus RNA-seq evidence.
5. Prepare outputs for downstream RNA-seq mapping and differential gene expression analyses.

Current assembly summary

The cleaned primary-contig assembly used for NCBI submission and EGAPx annotation has:

Metric	Value
Total assembly length	4,481,589,302 bp / 4.48 Gb
Number of contigs	4,081
Largest contig	36.84 Mb
Contig N50	6.88 Mb
N50 number	182
N90	629 kb
Gaps / Ns	0
GC content	43.07%

BUSCO, using vertebrata_odb10:

BUSCO category	Count	Percent
Complete BUSCOs	3,060	91.2%
Complete single-copy	2,952	88.0%
Complete duplicated	108	3.2%
Fragmented	134	4.0%
Missing	160	4.8%
Total searched	3,354	100%

BUSCO notation:

C:91.2%[S:88.0%,D:3.2%],F:4.0%,M:4.8%,n:3354

Repository structure

This repository currently contains scripts for several stages of the project.

Assembly

Script/file	Purpose
sparvus_hifiasm.sh	Run hifiasm assembly workflow.
sparvus_config.sh	Shared project paths/configuration variables for assembly scripts.
final_stats.sh	Generate final assembly statistics.
GenomeAssemblyInteractiveDump.ipynb	Interactive notebook used during assembly exploration/QC.

Assembly QC

Script/file	Purpose
2_sparvus_busco.sh	Run BUSCO on the assembly.
run_gc_coverage_check.sh	Run GC/coverage screening to identify potential contaminants or outliers.
run_gc_coverage_subset.sh	Run GC/coverage checks on a subset.

NCBI contamination/adaptor screening

Script/file	Purpose
download_fcs_gx_db.sh	Download the NCBI FCS-GX database.
run_fcs_adaptor.sh	Run NCBI FCS adaptor screening.
run_fcs_gx.sh	Run NCBI FCS-GX contamination screening.
remove_contaminant.sh	Remove the contaminant contig identified by FCS-GX.

The final cleaned assembly removed one contaminant contig:

Contig: ptg003819l
Length: 19,494 bp
Assignment: nematode / Oscheius tipulae
Action: EXCLUDE

This represented approximately 0.0004% of the assembly.

EGAPx annotation

Script/file	Purpose
validate_inputs_annot.sh	Validate EGAPx input files before annotation.
run_sparvus_egapx.sbatch	Initial EGAPx SLURM run script.
resume_sparvus_egapx.sbatch	Resume EGAPx after a failed/interrupted run.
egapx_config/	Snapshot of EGAPx config files edited for Unity/SLURM.
egapx_instructions.md	Detailed step-by-step EGAPx setup and run notes.

The initial EGAPx run failed because run_mask_assm_stats ran out of memory. The run was successfully resumed after increasing memory for that process.

EGAPx version used:

EGAPx 0.5.2
EGAPx commit: 40ec7362576e4b93fa9c5bf0a6c400d9502b8a63
Support data version: egapxsupportdata_20251017

EGAPx annotation summary:

Feature category	Count
Total genes	41,598
Protein-coding genes	23,550
mRNAs	30,699
CDSs	30,699
Non-coding RNAs	14,814
Pseudo transcripts	3,691

In the GFF, the EGAPx total gene count corresponds to:

37,929 gene features
+ 3,669 pseudogene features
= 41,598 total gene/pseudogene features

Isoforms are not counted as separate genes in the GFF. They are represented as mRNA child features under parent gene features:

32,415 mRNA records
25,246 parent genes with at least one mRNA
4,231 genes with more than one mRNA isoform

Adult RNA-seq download for annotation rerun

Script/file	Purpose
download_GSE232203_adult_rnaseq.sbatch	Download adult male S. parvus RNA-seq data from SRA/GEO.
compress_GSE232203_existing_fastq.sbatch	Compress partially converted FASTQ files after fasterq-dump.

Adult RNA-seq source:

GEO: GSE232203
BioProject: PRJNA971214
Species: Staurois parvus
Samples: adult male brain, spinal cord, and leg muscle

These data may be added to the RNA-seq evidence for a future EGAPx rerun.

Conda environments

File	Purpose
assembly_env.yml	Assembly-related tools.
busco_env.yml	BUSCO environment.
annot_env.yml	Annotation-related tools.
detonate_env.yml	DETONATE/transcriptome QC environment.
sp_ont_polish.yml	ONT polishing environment.

EGAPx input structure

EGAPx was run using:

• cleaned genome assembly
• NCBI taxid for Staurois parvus: 386267
• paired Illumina RNA-seq reads
• tRNAscan enabled
• cmsearch enabled

Example EGAPx YAML:

genome: /scratch3/workspace/kfloer_smith_edu-simple/sparvus_hifiasm/assembly_output/sparvus_primary_contigs.ncbi.fa
taxid: 386267
short_reads: /scratch3/workspace/kfloer_smith_edu-simple/egapx_sparvus/input/short_reads.txt

trnascan:
  enabled: true

cmsearch:
  enabled: true

The short-read manifest is a two-column file:

sample_name /path/to/R1.fastq.gz
sample_name /path/to/R2.fastq.gz

See egapx_instructions.md for the full setup and run details.

Androgen receptor sanity check

Because androgen receptor is a gene of particular interest for this project, the EGAPx annotation was checked for AR.

EGAPx recovered one androgen receptor locus:

AR feature	EGAPx result
Contig	ptg000154l
Coordinates	1,491,873-1,803,334
Strand	+
Gene ID	gene-egapxtmp_016766
Locus tag	egapxtmp_016766
Transcript ID	rna-egapxtmp_016766-R1
Protein ID	egapxtmp_016766-P1
Exons	8
Predicted protein length	777 aa
Ortholog cross-reference	NCBIOrtholog:367

BLASTP of the predicted protein against anuran proteins returned top hits to androgen receptor proteins. The top hit was:

androgen receptor [Aquarana catesbeiana]
Query cover: 100%
E-value: 0.0
Percent identity: 92.86%
Hit length: 777 aa

This supports the EGAPx AR model as a likely near-complete/full-length androgen receptor ortholog.

Notes for reuse

Most scripts are specific to the Unity HPC environment and will need editing before use elsewhere.

Likely things to modify:

• SLURM partition/account/qos settings
• module names
• file paths
• Apptainer/Singularity cache and tmp directories
• memory/CPU/time values
• Nextflow configuration
• EGAPx config files

The scripts should be treated as reproducible project documentation and templates, not universal drop-in workflows.

Suggested workflow order

A rough order for using this repository is:

1. Configure project paths in sparvus_config.sh.
2. Run assembly with sparvus_hifiasm.sh.
3. Generate assembly statistics with final_stats.sh.
4. Run BUSCO with 2_sparvus_busco.sh.
5. Run GC/coverage checks.
6. Run NCBI FCS adaptor and FCS-GX contamination screening.
7. Remove contaminant contig if needed.
8. Validate cleaned assembly.
9. Prepare EGAPx input YAML and RNA-seq manifest.
10. Validate EGAPx inputs with validate_inputs_annot.sh.
11. Run EGAPx with run_sparvus_egapx.sbatch.
12. If needed, adjust EGAPx process resources and resume with resume_sparvus_egapx.sbatch.
13. Use EGAPx GFF/GTF/protein/transcript outputs for downstream RNA-seq mapping and DGE.