Lytechinus variegatus v. 3.0 Genome Assembly

README

	This README describes the files associated with the sea urchin Lytechinus variegatus 
	v. 3.0 genome assembly available for download via Echinobase's (www.echinobase.org) 
	file transfer protocol (FTP). The files include the assembly itself as well as gene
	models, protein models, and functional annotations of these models through three pre-
	existing databases: 1) Strongylocentrotus purpuratus v. 4.2 gene models ("SPUs");
	2) Uniprot Knowledgebase (UniProtKB) protein database; 3) NCBI RefSeq Invertebrate 
	protein database. 

CONTACT
	
	If you have questions about the assembly or datasets, please direct them to:
	
	Phillip Davidson, phillip.davidson@duke.edu
	Greg Wray, gwray@duke.edu

CITATION

	If you use these data for your own work, please cite the following publication:
	
	Phillip L Davidson, Haobing Guo, Lingyu Wang, Alejandro Berrio, He Zhang, Yue Chang, 
		Andrew L Soborowski, David R McClay, Guangyi Fan, Gregory A Wray, Chromosomal-Level 
		Genome Assembly of the Sea Urchin Lytechinus variegatus Substantially Improves 
		Functional Genomic Analyses, Genome Biology and Evolution, Volume 12, Issue 7, July 
		2020, Pages 1080–1086, https://doi.org/10.1093/gbe/evaa101
		
FILE DESCRIPTIONS

	The following files are available for download:
	
		1) Lvar_3.0_scaffolds.fasta.gz
			Scaffold-level assembly of the L. variegatus v. 3.0 genome
			
		2) Lvar_3.0_contigs.fasta.gz
			Contig-level assembly of the L. variegatus v. 3.0 genome
			
		3) Lvar_3.0_annotations_v1.0.tar.gz
			Directory containing the version 1.0 gene models, protein models, and model 
			annotations for the L. variegatus v 3.0 assembly. These are same models that 
			were included in the print-publication of the genome (see above) and were last
			updated May, 2020. This directory includes:
				
				a) Lvar_3.0_v1.0.gff
					GFF  file describing the feature locations of the gene models
				b) Lvar_3.0_v1.0.transcripts.noUTR.fasta
					FASTA file of the gene models with UTR sequence excluded
				c) Lvar_3.0_v1.0.transcripts.fasta
					FASTA file of the gene models with UTR sequence included when 
					available
				d) Lvar_3.0_v1.0.proteins.fasta
					FASTA file of the translated gene models
				e) Lvar_3.0_v1.0.topSPUhit.txt
					Gene model hits of the single best match to the S. purpuratus 
					v. 4.2 gene models via BLAST-P
				f) Lvar_3.0_v1.0.topUNIPROThit.txt
					Gene model hits of the single best match to the UniprotKB protein 
					models via BLAST-P
				g) Lvar_3.0_v1.0.topREFSEQhit.txt
					Gene model hits of the single best match to the RefSeq Invertebrate 
					protein models via BLAST-P
		
		4) Lvar_3.0_annotations_v2.0.tar.gz
			Directory containing the version 2.0 gene models, protein models, and model 
			annotations for the L. variegatus v 3.0 assembly. These models were last 
			updated July, 2020. This directory includes:
				
				a) Lvar_3.0_v2.0.gff
					GFF  file describing the feature locations of the gene models
				b) Lvar_3.0_v2.0.transcripts.noUTR.fasta
					FASTA file of the gene models with UTR sequence excluded
				c) Lvar_3.0_v2.0.transcripts.fasta
					FASTA file of the gene models with UTR sequence included when 
					available
				d) Lvar_3.0_v2.0.proteins.fasta
					FASTA file of the translated gene models
				e) Lvar_3.0_v2.0.topSPUhit.txt
					Gene model hits of the single best match to the S. purpuratus 
					v. 4.2 gene models via BLAST-P
				f) Lvar_3.0_v2.0.topUNIPROThit.txt
					Gene model hits of the single best match to the UniprotKB protein 
					models via BLAST-P
				g) Lvar_3.0_v2.0.topREFSEQhit.txt
					Gene model hits of the single best match to the RefSeq Invertebrate 
					protein models via BLAST-P
		
		5) README.txt
		
ASSEMBLY BASIC STATISTICS

	ASSEMBLY	
 		Assembly size 	         870.4 Mb 
 		# scaffolds 	           104 
 		N50 scaffold length 	  45.6 Mb 
 		Longest scaffold 	      96.7 Mb 
 		No. scaffolds >10 Mb 	    19 
 		No. contigs 	           466 
 		N50 contig length 	      5.85 Mb 
 		No. contigs >1 Mb 	       285 
 		N (%) 	                  0.02 
 		GC (%) 	                 36.31 

	BUSCO 	
 		Complete 	95.50%
 		Duplicated 	 0.60%
 		Fragmented 	 0.80%
 		Missing 	 3.40%
		
	ANNOTATIONS 
		VERSION 1.0	(from publication, 5/20)
 			# gene models	      27,232 
 			Average gene length 	12.6 kb 
 			% Start codon 	        93.4 
 			% Start and stop codon 	90.3 
		
		VERSION 2.0 (from 7/20 update)
			# gene models	      29,837 
 			Average gene length 	15.2 kb 
 			% Start codon 	        99.8 
 			% Start and stop codon 	99.6 

METHODOLOGY

	ASSEMBLY
		
		For a detailed description of the assembly process, please see the publication
		referenced above (Davidson et al. 2020, GBE, https://doi.org/10.1093/gbe/evaa101). 
		Briefly, PACBIO Long Read sequences were used to initially assemble the genome 
		at the contig-level using Canu (Koren et al. 2017). Redundant (diploid) regions of
		the genome were removed and the near-haploid level assembly was polished with 
		Pilon (Walker et al. 2014) for accuracy using 10x Genomics short-read sequencing 
		data. Lastly, Hi-C chromatin conformation sequencing data was used to assembly the
		contig-level assembly into chromosome-level scaffolds using HiC-Pro (Servant 
		et al. 2015), Juicer (Durand et al. 2016), and 3D-DNA (Dudchenko et al. 2017)
		
	ANNOTATIONS
	
		For a detailed description of the annotation process for the v 1.0 gene models, 
		please see the publication referenced above (Davidson et al. 2020, GBE, 
		https://doi.org/10.1093/gbe/evaa101). Briefly, a combination of Maker (Campbell et 
		al. 2014) and gene prediction tools including Augustus (Stanke et 
		al. 2006) and SNAP (Korf 2004) were utilized. These tools were informed by pre-
		existing RNAseq (Israel et al. 2016) and protein (S. purpuratus v. 5.0) datasets 
		(www.echinobase.org). Models were annotated with the three pre-existing protein 
		databases mentioned above using BLAST-P (Camacho et al. 2008): 
		1) Strongylocentrotus purpuratus v. 4.2 gene models ("SPUs"); 2) Uniprot 
		Knowledgebase (UniProtKB) protein database; 3) NCBI RefSeq Invertebrate protein 
		database.
		
		For the version 2.0 gene models from the July 2020 update, a different approach 
		for gene model assembly and annotation was taken. Briefly, pre-existing RNAseq 
		(Israel et al. 2016) and protein (S. purpuratus v. 5.0) datasets 
		(www.echinobase.org) were integrated into the BRAKER genome annotation pipeline 
		(Hoff et al. 2019). Next, gene models annotated as transposable elements were 
		filtered out. Finally, gene models were corrected and extended with Trinity-
		assembled transcripts (Grabherr et al. 2011) using PASA (Haas et 
		al. 2003). Then, as before, models were annotated with the three pre-existing 
		protein databases mentioned above using BLAST-P (Camacho et al. 2008): 
		1) Strongylocentrotus purpuratus v. 4.2 gene models ("SPUs"); 2) Uniprot 
		Knowledgebase (UniProtKB) protein database; 3) NCBI RefSeq Invertebrate protein 
		database.