MEMEX tries to answer the following questions:

1. "Which of my selected proteins have the best/worst chances to be cloned?"

2. "Which of my cloned proteins have the best/worst chances to be expressed?"

3. "Which of my expressed proteins have the best/worst chances to be soluble?"

The classification is based on the Naive Bayes approach exploiting differences between experimentally amenable and recalcitrant proteins from TargetDB. The classifiers have been trained and tested on integral membrane proteins with at least 3 transmembrane helices, therefore its application to other protein classes is not recommended.

The input protein sequences are categorized into two classes: YES - successful; NO - unsuccessful. Additionally, the probability of a class (from 0.5 to 1.0) is provided. The probability threshold value is set by default to 0.5. "By increasing it one can expect higher classification precision (selectivity) and lower recall (sensitivity). If the result cannot be calculated, a comment is written.

If you use this server, please cite:
Antonio J. Martin-Galiano, Pawel Smialowski and Dmitrij Frishman. Predicting experimental properties of integral membrane proteins by Naive Bayes approach. Proteins. 2008 Mar;70(4):1243-56.

Submit your sequence:

Input sequences
Please provide single protein sequence (only 20 amino acids symbols are allowed) of length between 60-2000 amino acids in multiple FASTA format

Sample of input sequence

According to experimental stage of interest choose one of the following classifiers to evaluate your protein: