ProteinCCD with construct scoring and ranking

The Protein Crystallographic Construct Design [1] ProteinCCD (https://xtal.nki.nl/ccd/) software we previously described in deliverable 7.2 aims to increase the efficiency of researchers producing soluble protein in amounts suitable for structural studies, by facilitating the design of several truncation constructs of a protein under investigation. ProteinCCD functions as a meta server that collects information from (web-based) external software that predicts from sequence secondary structure, disorder, coiled coils, transmembrane segments, domains and domain linkers. Viewing the protein sequence annotated with the prediction results allows users to interactively choose possible starts and ends for suitable protein constructs and designing primers needed for PCR amplification. ProteinCCD outputs a comprehensive view of all constructs and all primers needed for bookkeeping and/or ordering of the designed primers.
The functionality of ProteinCCD has been extended under 7.1 to a new computational platform allowing a more interactive and efficient interface to the user, and providing new analysis options. These include parallel processing of server requests, more efficient interface for construct design, more cloning methods, an extended collection of existing vectors, local execution of some algorithms for improving response time, new servers for meta-analysis, easy bookkeeping, and better data security.
Working towards the goal of this deliverable, to provide construct scoring and ranking we implemented several features to reach this goal. Automated alignments of the “work” protein to orthologues present in typical model species, are now provided to the user to facilitate better choices for constructs. All constructs can now be given to the users not only as the “native” protein sequence (as before) but also in the specific context of the cloning vector used for production, including purification tags, and the sequence after enzymatic cleavage of the tags. This is important, as each proteins version has different properties. The molecular weight, isoelectric point, and absorption coefficient for every construct is also computed, enabling the users to understand the properties of the produced proteins.
The final goal to rank the chances of successfully producing the proteins, is realized by assessing the chances to produce soluble proteins for each protein. A score from 0-1 is given by different servers, and provided to the users. The constructs can then be ranked according to these scores.