metapredict online (v2.4)

metapredict is a deep-learning based consensus predictor of intrinsic disorder and predicted structure.

metapredict.net offers a simple online portal to some of metapredict's features but for a single sequence. As of version 2.0 we have moved to the updated metapredict disorder engine.

To use paste a single sequence into the input box below and select one of plot disorder, get values, or get IDRs.

Generate a high-resolution, interactive, and download-able plot of the per-residue disorder only:

Generate a high-resolution, interactive, and download-able plot of the per-residue predicted AlphaFold2 structural confidence score only:

Generate a high-resolution, interactive, and download-able plot of the per-residue disorder and the predicted AlphaFold2 structural confidence score:

Downloads a comma-separated value file with the residue position, amino acid, and disorder score for easy plotting in your own software:

Identify the discrete IDRs and download as a .fasta file (with start end position in FASTA header):

Color my sequence (CIDER-like). CIDER is having some DNS issues, so if you want to color-code your sequences this feature is offered here:

only one sequence?

Yes - metaserve only supports single sequences at a time. If you want to analyze many sequences, there are two possible options:

You can download metapredict and use it locally. It's easy - click here to get started!
You can upload a multiple sequence file to our Google colab notebook: click here to get started

i want to predict conformational properties more!

Great! Head on over to our ALBATROSS Google Colab notebook for various options!

how to cite:

If you have used metapredict please consider citing our manuscript!

Metapredict: a fast, accurate, and easy-to-use predictor of consensus disorder and structure
Biophysical Journal, 120(20), 4312–4319.
Emenecker, R. J., Griffith, D., & Holehouse, A. S. (2021). Link to paper
In addition, we have a paper on bioRxiv that details changes made for the V2 update:

Metapredict V2: An update to metapredict, a fast, accurate, and easy-to-use predictor of consensus disorder and structure
BioRxiv 2022.06.06.494887
Emenecker, R. J., Griffith, D., & Holehouse, A. S. (2022). Link to paper
Finally, if you have used the predicted dimensions (Rg, Re, or normalized Rg and Re), please cite:

Direct prediction of intrinsically disordered protein conformational properties from sequence
bioRxiv 2023.05.08.539824
Lotthammer, J. M., Ginell, G. M., Griffith, D., Emenecker, R. J. & Holehouse, A. S. (2023) Link to paper
NB: As of March 15th 2022, the default disorder predictor for metapredict is has updated to 'version 2', a more accurate predictor that incorporates information from AlphaFold2. For the original metapredict implementation, please use the 'legacy' version, which is still available via the metapredict Python package.

about

metapredict was developed by Ryan Emenecker, Dan Griffith, and Alex Holehouse in the Holehouse lab.
Last updated August 27th 2022.

changelog (metapredict.net)

August 27th 2022

Improved stability and fixed bugs with non-standard amino acids.
Added link to the Google colab notebook for multiple sequence predictions online.
Updated with link to V2 preprint

March 15th 2022

Updated the back-end disorder prediction engine to the more accurate metapredict V2.
Improved robustness for sequence parsing, including clipping FASTA headers.
Stand-alone IDRs are now shaded on the visualizations, and the colored IDR sequences provided directly for easy of analysis.

May 8th 2023

Added predicted radius of gyration and end-to-end distance from ALBATROSS.

May 31st 2023

Updated the ALBATROSS prediction networks to version 2 (as per V3 preprint)

Oct 22 2023

(V2.4 update) Improved robustness for input parsing, add AFRC Rg and Re for each IDR, added length for each IDR