Atlantic_croaker_range

Repository for bioinformatics and genomic analysis of Atlantic croaker (Micropogonias undulatus) collected off the Eastern United States.

This repository provides the code for the third chapter of the doctoral dissertation Conservation Genomics of Marine Fish Populations: https://doi.org/doi:10.7282/t3-wgyk-5s02. This research was supported by a NOAA Margaret A. Davidson fellowship.

As genomic data is too large to be stored on GitHub, sequence data were uploaded to the NCBI SRA and metadata to GEOME-DB.

Metadata is available at: https://n2t.net/ark:/21547/GjB2.

PacBio Hifi Sequencing data for one individual (used to generate a de novo assembly): BioProject PRJNA1460334 https://www.ncbi.nlm.nih.gov/sra/PRJNA1460334.

Illumina reads (low-coverage whole genome sequencing) for 400 individuals: BioProject PRJNA1457870 https://www.ncbi.nlm.nih.gov/bioproject/1457870.

Genome assembly, pre-processing, mapping, and downstream analysis were run using Amarel, Rutgers University's high performance computing system. Statistical analysis was run using R v.4.4.1 on a MacBook Pro (Apple M3 Pro Chip with 36 GB memory).

De Novo Genome Assembly

Assembly using Hifiasm

Hifiasm v. 0.19.3-r572
sbatch hifiasm_code.sh

Evaluating Hifiasm assemblies

QUAST v. 5.2.0
sbatch quast_code.sh

Selected primary contig assembly to move forward based on highest N50, lowest L50, and a total length similar to other croaker species. N50 18,479,188; L50 17; GC% 42.08; # contigs 3921; total length 782,013,506.

BUSCO v. 5.4.6 using lineage actinopterygii on primary contig Hifiasm assembly (hifi4523.asm.bp.p_ctg.fa)
sbatch busco_code.sh
Complete:98.9%[Single:97.4%,Duplicate:1.5%],Fragemented:0.2%,Missing:0.9%,n:3640

BlobToolKit: identifying and filtering out contamination

Running Blastn as first step to identifying contamination (following https://blobtoolkit.genomehubs.org/install/#databases):
Blast v. 2.6.0
sbatch blastn_code.sh

sbatch blobtools_create.sh Plots show reads with GC content < 40% are not vertebrates. Will remove these reads.

Filtering out contamination with Blobtools:
sbatch blobtools_create_filter.sh

Filtered assembly: N50 19,238,659; L50 17, GC% 42.54; # of contigs 2860; total length 731,051,241

Moving forward with filtered assembly for mapping of Illumina reads: hifi4523.asm.bp.p_ctg.filtered.fa

Pre-processing and Mapping of Illumina reads

Scripts are adapted from Nina Therkildsen's lab's data-processing repository (https://github.com/therkildsen-lab/data-processing).

Reads were run through pre-processing steps in two batches (Run1 and Run2) based on sequencing runs and merged after mapping. See Therkildsen repository above for a description of the sample tables and lists used in the scripts.

Pre-processing

Adapter trimming

trimmomatic v. 0.39
Sample Lists/Tables: Sample_Table_Run1.tsv and Sample_List_Run1.tsv; Sample_Table_Run2.tsv and Sample_List_run2.tsv sbatch adapter_clipping.sh

Quality filtering

fastp v. 0.23.4 multiqc v. 1.14

Trimming to quality scores > 30
sbatch quality_filtering_q30.sh

One additional trimming step to remove index barcodes at the beginning of reads
sbatch trim2.sh

Mapping

bowtie2 v. 2.5.1
samtools v. 1.17

Build Bowtie Reference Index

sbatch build_bowtie_ref_index.sh
Outputs .bt2, .dict, and .fai files

Mapping with Bowtie2

Sample_List_Combined.tsv and Sample_Table_Combined.tsv
sbatch low_coverage_mapping.sh

Running samtools conversion to bam file
sbatch samtools_mapping.sh

Merging bam files

bam_list_merged.tsv
sbatch merge.bam.sh

Overlap clipping of merged bam files

bamutil v.
sbatch overlapclipping.sh

In-del realignment

GATK v. 3.7.0
sbatch realign_indels.sh

The resulting merged, overlap clipped, and realigned bam files are used as the inputs for ANGSD.

Downstream analysis using ANGSD

ANGSD v. 0.940
PCAngsd v. 0.98.2

SNP Calling with ANGSD

sbatch angsd_global_snp_calling.sh
Generates list of SNPs to be used for subsequent analyses.

Genotype-likelihood estimation

sbatch get_beagle.sh
Generates Beagle file: bam_list_realigned_mindp132_maxdp4000_minind0.beagle.gz

PCAngsd: PCA, Admixture, Selection Scans

Outputs from PCAngsd are used in the Statistical Analysis in R section

sbatch run_pcangsd.sh
Ouputs covariance matrix: pcangsd_bam_list_realigned_mindp132_maxdp4000_minind0.cov

Selection Scan
run_pcangsd_selection.sh
Outputs npy file: pcangsd_bam_list_realigned_mindp132_maxdp4000_minind0.selection.npy

Admixture with PCAngsd
sbatch run_pcangsd_admix.sh
Selects K=2 based on MAP analysis
Outputs: pcangsd_bam_list_realigned_mindp132_maxdp4000_minind0_e.admix.Q.npy, pcangsd_bam_list_realigned_mindp132_maxdp4000_minind0_e.cov