Portal Home

cleverMachine

cleverMachine

The clever machine is an algorithm to discriminate between two protein datasets using physico-chemical properties contained in their primary structures. The first dataset to submit contains a hypothetical signal (e.g., substrates of a particular chaperone) while the second represents the reference set (e.g., a cell lysate). This distinction is purely internal and only affects presentation of the results (there is no bias in the algorithm operation). Once the datasets are discriminated with a certain accuracy, a classifier is made available at service.tartaglialab.com/clever_classifier.



 

Submission reference: 921083


Run label (optional)



Email address (optional)



Run mode
Normal run is recommended for majority of users and provides simplicity and robustness. Switching to expert mode enables multitude of additional features aimed at advanced users but at potential cost of increased computational time. Read more what could be done.

Normal run
Expert mode


The expert mode allows submission of custom scales and enable cross-validating coverage score reporting. Please see documentation for detailed description.


Threshold calculation
Default is 0.75 and adaptive calculates multiple thresholds to find best dictimination. Selecting this option may prolong your submission.

Normal run - 0.75 threshold
Adaptive threshold


Information extraction
By default, signal is extracted from the whole protein sequence. Selecting the peak detection mode filters the signal to only take into account regions with strong signal.

Complete profiles
Peak detection


Smoothing window size

7 AA
10 AA
15 AA
31 AA
61 AA


Require password for model (optional)



Upload a Positive dataset (FASTA format, at least 10 and at most 104 sequences)


Upload a Reference dataset (FASTA format, at least 10 and at most 104 sequences)


Provide custom set of scales in format as defined




Submission status:

Send submission



Sample data

Just trying it out? Pre-populate the form with: (Please, check that you do not have active proxies!)

mRNA binding proteins - The human RNA-binding proteins
[Castello, A. et al. (2012) Insights into RNA Biology from an Atlas of Mammalian mRNA-Binding Proteins. Cell, 149, 1393–1406]