Tutorial: MITE entry creation

This tutorial gives an comprehensive overview about the MITE entry creation process. Detailed information about biological databases and reaction SMARTS creation can be found on their respective pages.

• Finding enzymes and papers
• Submitting an enzyme to MITE
• What happens after submission
• Reporting bugs
• Link collection

Finding enzymes and papers

Finding a tailoring enzyme to describe in MITE can be challenging. Here, a few steps and strategies are outlined

Important

Before you start working on an enzyme, check whether it's already in [MITE](MITE published entries) of if it is already in the submitted but non-reviewed entries. This can save you an afternoon.

Literature search is the best way to find papers about tailoring enzymes that would fit well into MITE. The go-to are review articles about classes of enzymes (e.g., halogenases). Additionally, the MITE paper collection is a good place to search for papers about tailoring enzymes.

Warning

Many papers do not report the UniProt ID or GenPept Accession of the enzyme that was investigated. This makes it impossible to create a MITE entry (except in cases where this accession can be inferred by other means, e.g. by writing the authors). The MITE paper collection designate such papers with insufficient information (Completion not possible).

Additional ways of finding papers are:

• Expasy: This database provides a list of enzyme classes and all of their enzymes with their corresponding UniProtKB/Swiss-Prot entries -> Expasy database

Submitting an enzyme to MITE

Nota bene: a video tutorial is available here.

Entry submission consists of:

• Find and cross-reference the protein accession IDs
• Start the submission
• Check whether you need an auxiliary enzyme
• Write the descriptions
• Drawing of the reaction and exporting SMILES & SMARTS
• Go through the checklist and submit

Find and cross-reference the protein accession IDs

Every enzyme in MITE must be identified by an accession number. These IDs link the entry to external protein databases, where every enzyme is uniquely identified.

Important

Every entry needs at least a UniProt or a GenPept ID. But always try to find both of them and cross-reference them.

Tip

For a primer on biological databases and their identifiers read Appendix: Biological databases

Step 1: Check the paper. Sometimes the IDs are in the paper or the supplementary information. Take whatever you find there and continue with step 2 to fill in the rest.

Step 2: Get the GenPept ID from MIBiG. MIBiG is a database of biosynthetic gene clusters (BGCs).

1. Use the filter in the repository to search for the main product of the pathway or the organism that produces it.
2. Check the Compound tab: is the main product the one described in your paper?
3. Check the Genes tab: find your enzyme. The GenPept ID is always listed there.

Step 3: Get the UniProt ID. Paste the GenPept ID into the UniProt search bar. Then decide which ID to use:

• UniProtKB entry is reviewed (gold star) and well annotated → use the UniProtKB ID
• otherwise → use the UniParc ID

Tip

This works the other way round, too: in a UniProt/UniParc entry, the GenPept ID is under Sequence → Protein sequence.

Start the submission

Entries are submitted via the MITE submission portal. Simply add your enzyme name in the form field, specify the UniProt ID or NCBI GenPept Accession, and click on Start Submission. Automated checks will make sure that the Accession ID exists and that it is not yet described in MITE.

Next, you will be forwarded to a submission page where there are several form fields to fill in. Some of these fields are mandatory, others are optional.

Enzyme information

Fill in general information about the enyzme. Additional cross-links can be provided, and organic and inorganic co-factors specified. At least one reference must be provided.

It is good practices to include the description (if applicable):

• the natural substrate
• the biosynthetic pathway the tailoring enzyme belongs to
• retired MIBiG or UniProt IDs
• non-substrates / non-products

Auxiliary Enzymes

Some enzymes need an auxiliary enzyme to function. These helper enzymes are essential for the reaction of the enzyme you're curating.

Example: flavin-dependent halogenases always rely on flavin reductases. But these are rarely encoded in the BGC, because they're usually recruited from primary metabolism. So in most cases, you don't add them.

Important

Only include the auxiliary enzyme if it is encoded in the biosynthetic gene cluster.

Reaction Information

Here, the hypothetical minimal required substructure and reaction information can be described. This information is represented as a reaction SMARTS. Please see the Tutorial: Reaction SMARTS for a step-by-step guide.

Additionally, cross-links to databases, tailoring reaction terms, and evidence qualifiers can be specified.

Checklist before submitting

• Checked that the enzyme isn't already in MITE
• All your accessions are crosslinked
• UniProtKB ID only if reviewed, otherwise UniParc
• Auxiliary enzymes only if encoded in the BGC
• Enzyme description includes natural substrate and pathway
• Retired IDs, non-substrates and stereochemistry issues noted in the description
• Reaction descriptions filled in

What happens after submission

After the form page, a preview allows to inspect the entry before submission. Make sure to check the reaction depiction as well as metadata for correctness. Once you hit submit, the entry will be deposited on the MITE Data GitHub repository, where it will be picked up and reviewed by one of our expert reviewers. Feel free to participate in the discussion.

Examples

Example for retired UniProt and Non-substrate

Single chlorination of carrier-bound S-isoleucine (L-Ile-S-KtzC) in the kutzneride pathway. UniProt A8CF92 retired. Non-substrate: allo-isoleucine. (MITE0000418)

Example for RDKit stereo chemistryerror

Hydroxylation of aminoacyl-S-carrier (3-chloro-β-tyrosyl-S-SgcC2) in the lidamycin pathway. Stereochemistry could not be depicted due to incompatibility with validation engine. Additional C-3-substituted substrates: iodine, bromine, chlorine, methyl, and non-substituted. (MITE0000414)

Example for non-substrate

Halogenation of ChlA precursor peptide in the chlorolassin pathway. Dichlorination is the most likely outcome. Terminal tryptophan (Pos: 14) is never chlorinated. (MITE0000406)

Reporting bugs

If you notice a bug or any other issue, please report it on the MITE GitHub:

1. Open a new issue.
2. Describe your problem in plain English.
3. If possible, add screenshots so the problem is easy to understand and reproduce.

Link collection

Link	Descrption
MITE submission	Submitting a new enzyme
MIBiG	Finding the gene cluster and the GenPept ID of your enzyme
UniProt	Finding the UniProtKB or UniParc ID
NCBI Protein	Finding GenPept IDs
NCBI BLAST	Finding your enzyme when the paper only gives a sequence
PubChem	Copying (partial) structures as SMILES instead of drawing them manually
Ketcher	Drawing your reaction and creating SMILES and SMARTS
Daylight SMARTS Theory	Documentation of the SMARTS language, for when you're unsure what a SMARTS expression means
Drytest SMARTS/SMILES reactions	Checks your SMILES & SMARTS reactions for validity

Author(s) of this page: Marcel Fröhlich, Susanna Kummer, Mitja M. Zdouc