Home
Softono
MLFoodProcessing

MLFoodProcessing

Open source Jupyter Notebook
GitHub
21
Stars
5
Forks
2
Issues
4
Watchers
1 year
Last Commit
Nutrition & Diet ML Frameworks

About MLFoodProcessing

MLFoodProcessing is a machine learning toolkit accompanying the research paper Machine Learning Prediction of the Degree of Food Processing by Menichetti, Ravandi, Mozaffarian, and Barabasi. It implements FoodProX and FPro, computational models that predict the degree of food processing for food items based on their nutritional and ingredient composition. The repository provides scripts for training, evaluating, and applying these models, enabling researchers to classify foods and analyze dietary datasets. Key use cases include nutritional epidemiology, food system research, and public health studies investigating the relationship between ultra-processed food consumption and health outcomes. The software combines Python, MATLAB, and R code, requiring Python 3.6.10 with packages such as scikit-learn, pandas, shap, networkx, and imbalanced-learn. A standard desktop with 16 GB RAM suffices for minimal use, while 64 GB RAM and 16 cores are recommended for optimal performance. Installation typically takes less tha

Platforms

Web Self-hosted

Languages

Jupyter Notebook

Links

Source Code
m
Published by
menicgiulia
Visit View Profile

README.md

View on GitHub

MLFoodProcessing

Paper: Machine Learning Prediction of the Degree of Food Processing

Authors: G. Menichetti, B. Ravandi, D. Mozaffarian, A-L. Barabasi

DOI: https://doi.org/10.1101/2021.05.22.21257615

License: AGPL v3

Overview of FoodProX and FPro

alt text

System Requirements

Hardware Requirements

All the codes provided require a standard computer with enough RAM to support the operations. For minimal performance, this will be a computer with about 16 GB of RAM and 4 cores. For optimal performance, we recommend a computer with the following specifications:

RAM: 64+ GB

CPU: 16+ cores

Software Requirements

The provided codes have been tested on the following systems:

macOS (12.4)

Windows 10

Software versions used:

  • Python: 3.6.10
  • MATLAB: 2022a
  • R: 3.4.0 (2020-10-10)

Installation Guide

In Python:

pip install pandas
pip install scipy
pip install matplotlib
pip install seaborn
pip install tqdm
pip install joblib
pip install shap
pip install rfpimp
pip install dynamicTreeCut
pip install networkx
pip install pyvis
pip install scikit-learn
pip install imbalanced-learn

Version Numbers for Python Packages

  • pandas: 1.1.5
  • scipy: 1.5.2
  • matplotlib: 3.3.4
  • tqdm: 4.63.0
  • joblib: 0.17.0
  • shap: 0.41.0
  • rfpimp: 1.3.7
  • dynamicTreeCut: 0.1.0
  • networkx: 2.5
  • pyvis: 0.3.1
  • scikit-learn: 0.24.2
  • imbalanced-learn: 0.8.1

//: # ()

R Environment Used in the EWAS Study

> sessionInfo()

R version 4.0.3 (2020-10-10)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 22621)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252 
[2] LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

attached base packages:
 [1] parallel  stats4    grid      stats     graphics  grDevices utils    
 [8] datasets  methods   base     

other attached packages:
 [1] DESeq2_1.28.1               SummarizedExperiment_1.18.2
 [3] DelayedArray_0.14.1         matrixStats_0.57.0         
 [5] Biobase_2.48.0              GenomicRanges_1.40.0       
 [7] GenomeInfoDb_1.24.2         IRanges_2.22.2             
 [9] S4Vectors_0.26.1            BiocGenerics_0.34.0        
[11] jsonlite_1.7.1              readxl_1.3.1               
[13] crayon_1.3.4                testthat_3.0.0             
[15] margins_0.3.23              mfx_1.2-2                  
[17] betareg_3.1-4               MASS_7.3-53                
[19] lmtest_0.9-38               zoo_1.8-8                  
[21] sandwich_3.0-0              survey_4.0                 
[23] Matrix_1.2-18               Hmisc_4.4-1                
[25] ggplot2_3.3.2               Formula_1.2-4              
[27] survival_3.2-7              lattice_0.20-41            
[29] car_3.0-10                  carData_3.0-4              
[31] foreach_1.5.1               dplyr_1.0.2                
[33] broom_0.7.2                 hash_2.2.6.1               
[35] IHW_1.16.0                 

loaded via a namespace (and not attached):
 [1] colorspace_2.0-0       ellipsis_0.3.1         modeltools_0.2-23     
 [4] rio_0.5.16             htmlTable_2.1.0        XVector_0.28.0        
 [7] base64enc_0.1-3        rstudioapi_0.13        bit64_4.0.5           
[10] flexmix_2.3-17         AnnotationDbi_1.50.3   codetools_0.2-16      
[13] splines_4.0.3          geneplotter_1.66.0     knitr_1.30            
[16] annotate_1.66.0        cluster_2.1.0          png_0.1-7             
[19] compiler_4.0.3         backports_1.2.0        htmltools_0.5.0       
[22] tools_4.0.3            gtable_0.3.0           glue_1.4.2            
[25] GenomeInfoDbData_1.2.3 Rcpp_1.0.5             slam_0.1-47           
[28] cellranger_1.1.0       vctrs_0.3.5            iterators_1.0.13      
[31] xfun_0.19              stringr_1.4.0          openxlsx_4.2.3        
[34] lifecycle_0.2.0        XML_3.99-0.5           zlibbioc_1.34.0       
[37] scales_1.1.1           hms_0.5.3              RColorBrewer_1.1-2    
[40] curl_4.3               memoise_1.1.0          gridExtra_2.3         
[43] rpart_4.1-15           latticeExtra_0.6-29    stringi_1.5.3         
[46] RSQLite_2.2.1          genefilter_1.70.0      checkmate_2.0.0       
[49] zip_2.1.1              BiocParallel_1.22.0    rlang_0.4.8           
[52] pkgconfig_2.0.3        bitops_1.0-6           lpsymphony_1.16.0     
[55] purrr_0.3.4            prediction_0.3.14      htmlwidgets_1.5.2     
[58] bit_4.0.4              tidyselect_1.1.0       magrittr_2.0.1        
[61] R6_2.5.0               generics_0.1.0         DBI_1.1.0             
[64] pillar_1.4.6           haven_2.3.1            foreign_0.8-80        
[67] withr_2.3.0            abind_1.4-5            RCurl_1.98-1.2        
[70] nnet_7.3-14            tibble_3.0.4           fdrtool_1.2.15        
[73] jpeg_0.1-8.1           locfit_1.5-9.4         data.table_1.13.2     
[76] blob_1.2.1             forcats_0.5.0          digest_0.6.27         
[79] xtable_1.8-4           tidyr_1.1.2            munsell_0.5.0         
[82] mitools_2.4

Typical install time on a normal desktop computer for all Python and R packages is less than 10 minutes.

License

This project is licensed under the GNU Affero General Public License v3.0.

Note: This code is provided for research and educational purposes only.
For inquiries about commercial licensing, please contact
Giulia Menichetti ‹[email protected]›.