Total kcat values
kcat to be predicted (80.12%)
| Metric | Value |
|---|---|
| Total of entries in CataPro input file | 2569 |
| Number of cofactor identified | 7 |
| Metric | Value |
|---|---|
| Entries with no catalytic enzyme identified | 22 |
| Entries with missing KEGG IDs | 546 |
| Entries with missing enzyme information | 327 |
| Metric | Value |
|---|---|
| Number of duplicates | 2047 |
Duplicates occur when multiple reactions share the same enzyme-substrate combination. A high number of duplicates may result from multiple enzyme complexes sharing the same catalytic enzyme.
| Metric | Value |
|---|---|
| Number of transport reactions | 2495 |
Transport reaction represents the transfer of one metabolite from one cellular compartment to another without altering their chemical structure. The enzymes catalysing these transport reactions are often not represented in experimental databases (BRENDA and SABIO-RK). We argue that kcat predictions for such reactions tend to be less reliable due to the limited availability of experimental data, which limits the performance of deep learning models. Consequently, these values should be interpreted with caution. The full list of the predicted transport reactions is available in the log file 'prediction2_YYYY-MM-DD_HH-MM-SS.log'.
This report provides the input needed to run the CataPro machine learning model (CataPro repository). Follow the instructions in the repository to set up the environment and generate kcat predictions.