Skip to main content

Data from: Opsins in Onychophora (velvet worms) suggest a single origin and subsequent diversification of visual pigments in arthropods

Cite this dataset

Hering, Lars et al. (2021). Data from: Opsins in Onychophora (velvet worms) suggest a single origin and subsequent diversification of visual pigments in arthropods [Dataset]. Dryad.


Multiple visual pigments, prerequisites for color vision, are found in arthropods, but the evolutionary origin of their diversity remains obscure. In this study, we explore the opsin genes in five distantly related species of Onychophora, using deep transcriptome sequencing and screening approaches. Surprisingly, our data reveal the presence of only one opsin gene (onychopsin) in each onychophoran species, and our behavioral experiments indicate a maximum sensitivity of onychopsin to blue–green light. In our phylogenetic analyses, the onychopsins represent the sister group to the monophyletic clade of visual r-opsins of arthropods. These results concur with phylogenomic support for the sister-group status of the Onychophora and Arthropoda and provide evidence for monochromatic vision in velvet worms and in the last common ancestor of Onychophora and Arthropoda. We conclude that the diversification of visual pigments and color vision evolved in arthropods, along with the evolution of compound eyes—one of the most sophisticated visual systems known.


Total RNA of five onychophoran species, including three species of Peripatopsidae (Euperipatoides rowelli, Phallocephale tallagandensis, and Ooperipatus hispidus) and two species of Peripatidae (Principapillatus hitoyensis [referred to as Epiperipatus cf. isthmicola in the published article], and Eoperipatus sp.) was extracted using TRIzol (Invitrogen, Carlsbad, CA) and the RNeasy MinElute Cleanup Kit (Qiagen, Hilden, Germany) according to the manufacturers protocols. Double-stranded complementary DNA was prepared following Illuminas messenger RNA (mRNA)-seq sample preparation guide (Part #1004898 Rev. D). Briefly, mRNA was isolated from total RNA using Dynal oligo(dT) beads (Invitrogen) and fragmented (~200bp) using divalent cationic ions. First-strand synthesis was performed with random hexamer primers, using Superscript II polymerase (Invitrogen). Subsequent second-strand synthesis was carried out with DNA Pol I polymerase (Invitrogen). Starting at the blunt end repair step, multiplex sequencing libraries were prepared following the multiplex protocol of Meyer and Kircher (2010), with modifications for double indexing described in Kircher et al. (2012). These libraries were sequenced according to the manufacturers instructions with 75 cycles paired end on the Genome Analyzer IIx platform (v4 sequencing chemistry and v4 cluster generation kit; Illumina, San Diego, CA).

Raw sequences were analyzed with IBIS 1.1.2 (Kircher et al. 2009). For highly accurate sample identification, sequences with falsely paired indices were discarded. Paired-end reads from a single cluster were merged, if at least 11 bp were overlapping (Kircher et al. 2011). Reads with low complexity were removed. For each library from the five species, we trimmed the Illumina raw data at three different levels of stringency (Filter15: reads with more than five bases below a phred quality score of 15 were removed; Filter30: reads with less than 20 bases in a row with a phred quality score of at least 30 were removed; Filter25: same as Filter30 but the threshold value was set to 25 to keep as many high-quality reads as possible, 63–89% compared with Filter15). This was done to deal with the tradeoff between the quality and quantity of reads as we lost many reads (51–90% compared with Filter15) by using the very stringent Filter30, which thus might have led to a loss of lowly expressed genes (for details see statistics_and_BUSCO_values.xls). For all libraries and filters described, sequence reads were assembled de novo using the CLC Genomics Workbench 5.1 (CLC bio, Århus, Denmark) with the following parameters: mismatch cost 3; insertion cost 3; deletion cost 3; length fraction 0.5; similarity fraction 0.8; minimum contig length 200; automatic word size; automatic bubble size; and contig adjustment by mapped reads (for details see statistics_and_BUSCO_values.xls).

Usage notes

An explanation of each file is given in the README.txt file.