De novo transcriptome and differential expression in body regions of Leucosolenia corallorrhiza (Porifera, Calcarea)
Description
Supplementary data for the publication Bolshakov et al. Structure, transcriptional signature, and regeneration of oscular rim in calcareous sponge Leucosolenia corallorrhiza // Invertebrate Zoology. The data include: - de novo transcriptome (leuc_95.fasta); - transcriptome annotation (full_annotation_final.xlsx); - data matrix with transcript counts at the “isoform level” for body regions (Data_matrix_isoforms.rar); - data matrix with transcript counts at the “gene level” for body regions (Data_matrix_genes.rar).
Files
Steps to reproduce
For RNA sequencing, four large sponges L. corallorrhiza were split into body regions: cormus tubes (C), diverticula (D), and oscular rims (O). Additionally, one large sponge was processed completely. The samples were lysed in 1 ml of ExtractRNA (Evrogen, BC032), and RNA was extracted according to the manufacturer's protocol. The polyadenylated fraction of total RNA was purified with NEBNext Poly(A) mRNA Magnetic Isolation Module. Libraries were constructed with the NEBNext Ultra II RNA Library Prep Kit and sequenced on the Illumina NovaSeq6000 in paired-end mode with 150 bp read length. A de novo transcriptome was assembled by Trinity (v2.15.1) based on pooled paired-end reads from all sampled points (nine experimental and one reference samples) with default parameters and in silico read normalization. For differential expression analysis, cleaned paired-end reads from nine body region samples were aligned to the assembled reference transcriptome using Bowtie2 (v2.5.4), and transcript abundance was estimated with RSEM (v1.3.3) with parameters implemented into Trinity pipeline scripts. Differential expression analysis was performed in R (v4.5.1) using DESeq2 package (v1.48.2) at the level of Trinity’s “genes”. Unnormalized transcript-level count data were imported into the DESeq2 using the tximport package (v1.36.1). A minimal pre-filtering step was applied following the DESeq2 recommended workflow, retaining only transcripts with at least 10 counts in a minimum of three samples. Exploratory data analyses revealed unwanted variation resulting in samples being grouped by the individual from which they were obtained (rather than by body region). This fact was accounted for by including a batch variable in the DESeq2 design formula (~individual+condition). For downstream analyses and data visualization, the batch effect of variance-stabilized transformed (vst) counts was further corrected using the limma package (v3.64.3). Batch effect correction resulted in proper sample clustering corresponding to the studied body regions. Differential expression testing was carried out using pairwise comparisons between body regions (O vs C, O vs D, and D vs C). Additionally, we compared O vs the combined C+D condition by specifying numeric contrasts manually. For additional information on sample preparation and data analysis, see Bolshakov et al. Structure, transcriptional signature and regeneration of oscular rim in calcareous sponge Leucosolenia corallorrhiza // Invertebrate Zoology.
Institutions
- Lomonosov Moscow State UniversityMoscow, Moscow