ProteinPred is a computing tool designed to predict biochemical, physical-chemical, and nutritional properties
of individual proteins or protein groups. This analysis is based on their amino acid sequences or experimental amino acid compositions.
ProteinPred provides three types of user inputs:
Calculated parameters includes :
- Amino Acid Compositions & Chemical Scores - Biophysical Chemical Parameters : molecular weight, isoelectric point, hydrophobicity, fraction of charged residues, net of charge per residue,
partial specific volume, protein volume, extinction coefficients, nitrogen-to-protein conversion factor, partial specific volume, refractive index increment (dn/dc), protein volume
- Theoretical UV-Visible Spectra,
- Wavelength Dependence of dn/dc,
- Protein Charge & capacitance as a function of pH,
- Theoretical Sorption Isotherm and related parameters,
- Scattering-Related Parameters: atomic composition, scattering length density (SLD), D2O/H2O matching point & electron density.
Explore your results on the web interface or download them as CSV files in a ZIP archive or as an Excel workbook.
Visit the Documentation tab for detailed explanations of calculations, corresponding references for citation,
and contact information for any questions or comments.
This application was developed and maintained by INRAE Transform Plastic.
Download the template CSV filefor analysing a group of protein sequences. Fill it with your data and upload it below.
You will find the guidelines to fill the CSV file in the documentation tab. You can also test the app by downloading the following example of CSV file.
Atomic composition table Scattering related parameters
Upload a CSV File
Download the template CSV filefor analysing a group of experimental amino acid compositions. Fill it with your data and upload it below.
You will find the guidelines to fill the CSV file in the documentation tab. You can also test the app by downloading the following example of CSV file.
Atomic composition table Scattering related parameters
FoodProteinsDB Explorer
Explore molecular descriptor diversity across food proteins
Click a point for protein information. Select rows in the table for export.
Descriptor distribution
Amino Acid Profiler
Compare the amino acid composition of a protein or protein group with a reference of your choice.
Choose a target.
Select filtered food proteins, a protein group, a family, a food source, or My ProteinPred result. For your results, choose the calculation mode in the left panel, then one protein or all calculated proteins.
Choose a reference.
Use the whole database or a specific group, such as RuBisCO. Only Filtered FoodProteinsDB follows the filters on the left.
Read the profile.
Relative composition shows enrichment above zero and depletion below zero. Residue composition (%) compares the target bars with the reference line. For several target proteins, bars show their mean composition; ±1 SD optionally shows their spread.
Relative composition = (target mean − reference mean) / reference mean. +1 means twice the reference proportion; −1 means absence. A zero reference proportion is undefined and omitted. Each protein contributes equally. Profiler groups use explicit family names; the dairy/egg cohort excludes caseins. SD describes variation among target proteins, not statistical confidence; in relative mode it is divided by the reference mean, treated as fixed.
Residue groups reproduce the supplied profiler: disorder promoting (A R S Q E G K P), order promoting (N C I L F W Y V), neutral (D H M T). These compositional tendencies do not establish a protein's disorder.
FoodProteinsDB provides a reference for molecular descriptor diversity across food proteins. Your ProteinPred results can be compared with this diversity. An atypical position may reflect a particular composition or motivate experimental measurements; it does not establish functionality.
Reference: archived FoodProteinsDB calculations from the supplied 2022 workbook. User results: current ProteinPred calculations. dn/dc values correspond to the compositional calculation at 589.3 nm and 25 °C (Zhao, Brown & Schuck, 2011). Absolute NCPR and net charge at pH 7 are distinct descriptors. Chemical score uses the FAO 2011 pattern and does not account for digestibility.
Welcome to Help Section
User input
(1) Single sequence module
You need to fill the following input fields:
Protein name: no specific requirements,
Amino acid sequence: unlimited number of letters comprised in the natural 20 amino acids code (fasta format accepted),
Number of polypeptide chains: numerical value, not exceeding the total number of amino acid
Number of disulfide bonds: numerical value, not exceeding twice the cystein content.
Click on the "Load Example" button for demonstration purposes.Click on the "Reset" button to remove all the fields.
(2) Multiple sequence module
The CSV file to updload need to fulfill the following requirements:
Contain at least 4 columns with the column names provided in the template,
Contain at least 1 row,
Protein name: unique value,
Amino acid sequence: unlimited number of letters comprised in the natural 20 amino acids code (fasta format not accepted),
Number of polypeptide chains: numerical values, not exceeding the total number of amino acid
Number of disulfide bonds: numerical values, not exceeding twice the cystein content.
The CSV file to updload need to fulfill a following requirements:
Contain at least 24 columns with the column names provided in the template,
Contain at least 1 row,
Protein name: unique value,
Number of polypeptide chains: numerical values, not exceeding the total number of amino acid
Number of disulfide bonds: numerical values, not exceeding twice the cystein content.
Amino acid composition: expressed in g/kg of powder for 18 amino acids ,
Molecular weight: expressed in g/mol,
The module 'Amino acid Composition' requires information beyond amino acid composition.
Specifically, experimentally, asparagines and glutamines are analyzed in their acidic forms.
Therefore, an estimation of the ratios of asparagine (N) to aspartic acid (D) and glutamine (Q) to glutamic acid (E) is needed.
These ratios are expressed as follows:
Ratio_N_N+D = weight of N / (weight of N + weight of D),
Ratio_Q_Q+E = weight of Q / (weight of Q + weight of E),
Weights are expressed in mg/g of protein. If these ratios are unknown to the user, they can be determined using modules 1 and 2 from the sequences.
Additionally, theFoodProteinsDBdatabase lists the amino acid composition and associated molecular weights for many food proteins.
You can proceed by clicking on the "Launch analysis" button. The analysis is computed.
You can visualize your data on the bottom part of the app.
You can get the indivual data either in each tabset using the "Copy", "Download", and/or "Save plot" buttons
All the data may be downloaded or sent by email on the top right panel.
About ProteinPred
Applicability of ProteinPred
ProteinPred is a computing tool designed to meet the needs of researchers, engineers, and students in Food Science & Technology .
It may also be of interest to broader communities in biophysics, biotechnology, or biomedical fields.
People behind ProteinPred
The calculation, iconography, and design behind ProteinPred were built by Adeline Boire, research scientist at INRAE,
BIA Nantes. The ShinyApp and the informatic infrastructure behind to host the website were developed by Michard Rakotoson,
computer engineer at INRAE, Plastic (TRANSFORM platform).
For bug reports and feature requests, please contact Adeline Boire at adeline.boire@inrae.fr .
The following people have contributed to the development of ProteinPred: Maude Ducrocq, Aurélie Nicolas, Stéphane Pezennec, Antoine Bouchoux, Denis Renard and Hamza Mameri.
Fundings
The application was developped in the framework of the project FoodProteinsDB, founded by
INRAE TRANSFORM department department during the period of 2022-2023.
Reference
Paper in preparation.
Version history
v1.1.0 — 2026
Major performance improvements for large batch analyses.
Introduction of the FoodProteinsDB Explorer for contextualizing predicted protein properties.
Faster rendering of tables, figures and interactive plots.
CSV and Excel export available in parallel.
No change to the underlying scientific equations, constants or prediction models.
v1.0.0 — 2024
Initial release of ProteinPred.
Calculation of sequence-derived physicochemical and biophysical descriptors.
Support for single-protein and batch analyses.
Export of calculated properties and associated data.
