Frequently Asked Questions
How can I access the data?
Our website is located at our GitHub repository and all the data is freely available.
How were conservation scores calculated?
We calculate conservation percentage by removing gaps and applying a similarity measure. For each position, we first calculate the gap percentage. If the gaps are the most frequent, we assign 0% conservation to the corresponding position. If not we count the most frequent amino acid and add other amino acid(s) that give a BLOSUM80 score greater than 1. The count of similar amino acids are divided by the number of non-gap sequences.
How were divergent orthologs identified?
Divergent sequences were manually identified by examining features such as missing transmembrane regions and regions of extreme or unexpected sequence divergence. N- and C-terminal regions were mainly excluded from this analysis. We welcome community feedback to help improve the divergent ortholog set — if you have suggestions or corrections, please reach out to us.
What is the header format for protein sequences?
sp: Swiss-Prot / tr: TrEMBL|Uniprot Protein ID|Gene Name _ Species Name|Tax ID
How were sequences assigned to GPCR families?
The members of each family were taken as one or more clades of its search tree, after ordering the tree by search score. They were then kept only with a single GPCR-like 7TM core and a topology score of at least 0.8, and sequences claimed by two families were assigned to one of them. The Family Assignment page shows the search tree, the tree-ordered alignment, the search scores and the extracted clade of every family after each of these steps.
How was the topology of the sequences used to filter them?
Every residue of each family's alignment was replaced by its predicted topology state, and each sequence was scored by how well its transmembrane helices line up with the seven consensus helices of its family. Sequences scoring below 0.8 were removed. The Topology Filter page shows how many sequences the filter removed from each family and the topology alignment of every family, with the removed sequences marked.