Kinase expression constructs

Information

Background:

The aim is to select a panel of kinases suitable for E. coli expression in 96-well plates.
Ideally, the kinase catalytic domain is to be expressed in isolation.
Plasmids are provided from two libraries:

• SGC Oxford
• Dana-Farber/Harvard Cancer Center (DF/HCC) Harvard Institute of Proteomics (HIP) human kinase collection (pJP1520; MMLV retroviral, puromycin resistance vector, complete set)

The PDB is used here as a source of constructs which have been successfully expressed. For each kinase target, the PDB is searched for matching constructs from any species, expressed in E. coli.

Custom selection procedure:

• All kinase domain targets with a matching SGC Oxford plasmid suitable for E. coli expression.
• Plus all other kinase domain targets which have a matching HIP pJP1520 plasmid, as well as at least one matching PDB construct from any species, expressed in E. coli, and with an authenticity score of > 0.
• The authenticity score is an attempt to circumvent problems assocaited with the frequent misannotation of sequences in PDB entries (the SEQRES records often contain only the structurally resolved sequence instead of the actual sequence used in the experiment). The score is generated by searching for expression tags within a PDB construct—anything with a detected expression tag is given a positive authenticity score. The assumption here is that expression tags are generally not resolved, and thus SEQRES records containing expression tags are less likely to have been erroneously annotated with only the observed sequence.
• If multiple suitable constructs are available for a kinase domain target, the construct with the lowest number of extraneous residues outside the catalytic domain region will be selected for expression testing.
• Selected PDB construct sequences will be mapped onto matching plasmid sequences and subcloned.

In formal notation:

( plasmid_source = "SGC Oxford") or ( nPDBs > 0 and top_pdb_auth_score > 0 )

Description of table fields with details of selection procedure:

targetID	Kinase domain targets are given an identifier consisting of UniProt entry name followed by the kinase domain index (since some kinases have two catalytic domains)
DB_target_rank	Qualitative ranking of human kinase importance, based on metrics such as number of bioassays, number of PDBs, number of publications, % mutation rate in sequenced tumor samples
plasmid_ID	For a given kinase target, all matching plasmids from the available plasmid libraries (see plasmid_source) are aligned and sorted based on nconflicts and nextraneous_residues (see below), and the top-ranked plasmid is selected (click the targetID to see the alignment)
plasmid_source	Source of the top-ranked plasmid. Two plasmid libraries are currently available to the Chodera lab - SGC Oxford and HIP pJP1520. Only SGC Oxford plasmids suitable for E. coli expression are included in this selection procedure.
plasmid_nconflicts	Number of conflicts between plasmid and UniProt sequence in the region of the target domain
plasmid_nextraneous_residues	Number of plasmid residues outside the target domain region
nPDBs	Number of PDB constructs matching the target gene, from any species, and with an E. coli expression system annotation
top_pdb_ID	Matching PDB constructs are ranked by auth_score (see below), nextraneous_residues, and nconflicts, and the top-ranked PDB (and chain ID) is selected
top_pdb_expr_tag	Type of expression tag detected for top-ranked PDB construct
top_pdb_auth_score	Authenticity score assigned to top-ranked PDB construct based on the expression tag detected. Currently this is only used to distinguish between constructs with expression tags (auth_score > 0) and those without (auth_score = 0)
top_pdb_nextraneous_residues	Number of PDB construct residues outside the target domain region
family	UniProt protein family annotation
top_pdb_taxname	Source organism of the top-ranked PDB construct
selected_construct_source	Selected constructs will ultimately be full-length SGC Oxford plasmids or PDB constructs (to be subcloned from either SGC Oxford or HIP plasmids).
selected_construct_nextraneous_residues	nextraneous_residues for the selected construct

Other notes:

Sort table by clicking on column headers. Shift-click for multicolumn sorting. Drag-click to rearrange columns. Refresh page to reset.
Click on the targetID to see alignments of plasmids and PDB constructs against the target UniProt sequence (kinase catalytic domain highlighted in red).