nenMC - Predict Organellar Location of Proteins
nenMC consists of two neuronal networks, nenM and nenC, which predict the mitochondrial and chloroplast localization of proteins, respectively.
nenM
evaluates the first 25 amino acids and currently possesses 100 input, 4 hidden and 2 output neurons.
nenC
evaluates the first 60 amino acids and currently possesses 300 input, 1 hidden and 2 output neuron(s).
Both networks were generated with Predotar's traing set.
Please note that this is a pre-liminary version of the application and currently we do
no format checks, no sequence checks; format error -> no result.
Thus, please make sure to submit:
- genuine FASTA format (single or multi)
- >= 60 aa per sequence (no "X" allowed, but case insensitive)
Confidence values are still under evaluation but rules of thumb are:
| Organellar localization |
if Confidence chloroplast (C = 1) |
if Confidence mitochondrial (M = 1) |
| > 95% | > 128 | > 170 |
| > 90% | > 74 | > 150 |
| > 80% | > 30 | > 85 |
| possibly | > 0 | > 0 |