Assignment #10 - Building your own R package

This week, our objectives were:
- Learn the structure of an R package and the role of the DESCRIPTION file
- Practice writing machine-readable metadata (authors, dependencies, versioning)
- Draft a coherent package proposal for your final project
- Publish your package scaffold proposal online (GitHub and blog)

My R package: disordR
  • Purpose and scope of the project
    • The disordR package is a small toolkit designed to be used for students or researchers studying intrinsically disordered proteins (IDPs). It has a simple pipeline that can be used in class projects or advanced research. First, it will calculate charge-hydropathy metrics (based on the Uversky plot). It will then create a concise output of disorder predictors, and finally, it will map AlphaFold pLDDT to assumed intrinsically disordered regions (IDRs). 
  • Key functions I plan to implement
    • aa_props() - mean hydropathy, net charge, and fraction of charged residues
    • uversky_metrics() - hydropathy and charge per residue; classifies IDP or ordered
    • uversky_plot() - scatter plot vs Uversky boundaries
    • consensus_disorder() - mean consensus or predictor scores
    • idr_segments() - IDR intervals
    • read_af2() - per residue pLDDT from AlphaFold
    • idr_from_plddt() - IDR segments from pLDDT
  • How I chose the fields in DESCRIPTION
    • Package/Title/Description: I wanted to keep the description simple and easy to understand by highlighting the three main functions (charge-hydropathy, consensus disorder, pLDDT IDRs)
    • Version: 0.0.0.9000 is the development version provided by the professor for the class project
    • Authors@R: I am listed as the author for accountability, contact, and rights.
    • Depends: R (>= 3.1.2) as listed by the professor
    • Imports: tibble, dpylr, stringr, readr, ggplot2 for clean data management and simple plot generation. bio3d is listed under "Suggests" due to potential issues with AlphaFold and extended usage or run time. Therefore, the package can still build even if bio3d is not installed.
    • License: CC0, since it makes reuse simple
    • Encoding: UTF-8, avoids issues with dashes/quotes in metadata
    • LazyData: "true" since a small dataset may be included
  • Link to GitHub repo for disordR

Comments

Post a Comment

Popular posts from this blog

Assignment #5: Matrix Algebra in R

Image
  R code for creating A and B, and for computing  invA ,   detA ,   invB , and   detB : Outputs for each operation: Why  solve(A)  and  det(A)  behave this way: Since Matrix A is 10x10 and therefore a square, it could technically have an inverse. However, since the determinant of A is 0, it is considered a singular matrix, meaning it does not have full rank and therefore cannot be inverted. Due to this,  solve(A)   returns an error even though the matrix is a square, yet it is still able to run and confirm the singularity. Why operations on B fail (non‑square matrix): Since Matrix B is 10x100, it is not square. Inverses and determinants can only be defined for square matrices, meaning both solve(B) and det(B) return errors. Though it is expected, it highlights how important it is to check matrix dimensions before attempting further operations.  Notes on numeric stability or performance: There are important considerations w...
Read more

Final Project: disordR

  disordR: a minimal IDP toolkit in R Project Goal  I built disordR to provide a small toolkit for analyzing intrinsically disordered proteins (IDPs). The package includes functions for amino-acid properties, classic Uversky charge-hydropathy metrics/plots, and a simple consensus combiner for per-residue disorder scores. A small, bundled dataset makes testing the functions simple and reproducible. Design Reasoning I chose four  functions to cover common IDP tasks: aa_props() → mean Kyte–Doolittle hydropathy, net charge, and fraction charged residues (FCR). uversky_metrics() → scaled hydropathy (0–1) + mean net charge per residue with a simple IDP vs Ordered call uversky_plot() → classic Uversky scatter with the boundary line for visual interpretation. consensus_disorder() → mean consensus or predictor scores I limited dependencies to tibble and ggplot2 to make installs reliable and simple. A f uture direction of this package could include lig...
Read more

Assignment #9 - Visualization in R – Base Graphics, Lattice, and ggplot2

Image
R code: GitHub link Generated plots:  Scatter plot:  Histogram: Conditional scatter plot (lattice): Scatter with linear trend (ggplot2): Discussion:  How do the syntax and workflow differ between base, lattice, and ggplot2? Base R is very simple and step-by-step, where you first call 'plot( )', and then begin to add what you want line by line, such as points, lines, or legends. With lattice, you establish the description of the plot once with a formula 'xyplot(y ~ x | group, data=df)', with optional formatting lines, and then it renders the plot from that. It is more rigid with edits and requires you to edit the actual code rather than additional lines. Finally, ggplot2 is built together, data + aesthetics + layers 'ggplot(df, aes(x, y, color=group)) + geom_point( ) + geom_smooth( )', therefore allowing it to be easily tweaked.  Which system gave you the most control or produced the most “publication‑quality” output with minimal code? I would say that ggplot2 pr...
Read more