snaketron / clustirr Goto Github PK

• tutorial how to use it

gliphR vs. Jan's turboGliph vs. gliph's original versions vs. ting

Benchmarking to show
a) consistency: we should hope to find similar results given a simple sample
b) that choices in gliphR are biologically more suitable by examining clonaly expanded samples

on b) do we have an appropriate naive reference?

• At first for us against ground truth for sanity checks
• later on maybe included vignette in follow up version

The original gliph algorithm uses as input the following:

• minimum: vector of CDR3b sequences
• maximum: data.frame with CDR3b + V + J (+ 3 columns for alpha chain)

To use V+J information in such a way that processes of local/global clustering are affected, the
user also has to use setup additional input parameters. It is my impression that very few users
do this, i.e. most users will provide CDR3b sequences only as input.

Hence my suggestion for gliphR:

We use as main input (parameter data_sample) a data.frame with 1 or 2 columns:

• if 1 column -> the column has to be named CDR3a or CDR3b
• if 2 columns -> the columns will represent CDR3a and CDR3b (order not relevant)

What do you think?

This is a general TODO list

• input checks

• #12

• #13

• #14

• #15

• #18

• #19

• DESCRIPTION + NAMESPACE update

• maybe we need to use some interfaces provided by Bioconductor

In some of our earlier meetings we talked about the visualization of ClustIRR. This is a summary:

• expand_clones parameter is not needed. This makes the algorithm computation/memory-wise more tractable
• plot_graph.Rd -> graph.Rd; it is more convenient to name the man files the same as the scripts they describe. In my opinion, we need to work a little bit more on the text in this man page. Novice user should be able to use this page as a reference to interpret all visual symbols (shapes, colors, etc.) they encounter in the output.

If we remove expand_clones we can also simplify the vignette.

• replace or complement get_edges() function with function to receive igraph compatible data object

https://github.com/snaketron/ClustIRR/blob/531e13aa1969e4b94f6e432d0f3457809df00c98/R/graph.R#L120C11-L120C32

@kaozkai

I though we were creating undirected (simple) graph. Is this correct?

TODO:

• finish section about edges
• transfer left parts about version 1 and version 2 into the comparison vignette
• add section with minimal graphical output

Internally global distances are computed between CDR3 sequences using hamming distance.

Well-known tools, such as tcrdist3, already exist that can compute global distances using more sophisticated. We should at least provide an interface to skip global dist computation and to use the distances estimated by complementary methods.

Single-cell RNA sequencing coupled to TCR
profiling of large granular lymphocyte leukemia
T cells

Data with HLA included

Find out were the noise/scattering comes from in OvE values when identical input is put in (possibly related to sample replacement)

@kaozkai Which functions should be exported by ClustIRR?

Currently, we export:

export(cluster_irr)
export(get_graph)
export(get_edges)
export(plot_graph)

I would guess we need cluster_irr, get_graph and plot_graph.
Do we really need to have get_edges visible to the user?

For get_graphs

General ToDo for submission process

• add comparison plots
• add section with minimal graphical output

Update gliphR with additional features after final release

Future ToDo List, for example scoring

General procedure:

1. make gliphR Bioconductor conform (test, input checks, vignettes, document, ...)
2. submit clustering algorithm to Bioconductor
3. update gliphR with scoring (quantification part)

Incorporate all changes once the package is submitted

Future ToDo List

In function R/input_checks we perform parameter checks.

We can borrow Jan's checks, but rename the parameters.

Additional checks have to be done for missing parameter
values, NAs, or NULLs, etc.

Errors

• package size: Package tarball exceeds the Bioconductor size requirement. is: 6.2mb should be: <5mb (added hs_CD4_ref.RData to build ignore for now)
• valid maintainer: Remove Maintainer field. Use Authors@R [cre] designation.
• vignette directory: No 'vignettes' directory. (created dummy vignette)
• man page documentation: At least 80% of man pages documenting exported objects must have runnable examples. (added minimal example, to be expanded with meaningful one)

Warnings

• DESCRIPTION/NAMESPACE consistency: Import methods in NAMESPACE as well as DESCRIPTION.
• coding practice: Avoid T/F variables; If logical, use TRUE/FALSE

Notes

• Description: field: The Description field in the DESCRIPTION is made up by less than 3 sentences. Please consider expanding this field, and structure it as a full paragraph
• coding practice:
  • Avoid 1:...; use seq_len() or seq_along()
  • Avoid the use of 'paste' in condition signals
• function lengths: The recommended function length is 50 lines or less. There are 5 functions greater than 50 lines.
• unit tests: Consider adding unit tests. We strongly encourage them. See https://contributions.bioconductor.org/tests.html
• formatting of DESCRIPTION, NAMESPACE, man pages, R source, and vignette source:
  • Consider shorter lines; 4 lines (0%) are > 80 characters long. (roxgen auto-created, man/gliph.Rd)
  • Consider multiples of 4 spaces for line indents; 102 lines (5%) are not. See https://contributions.bioconductor.org/r-code.html / See styler package: https://cran.r-project.org/package=styler as described in the BiocCheck vignette.
• bioc-devel mailing list subscription: Cannot determine whether maintainer is subscribed to the Bioc-Devel mailing list (requires admin credentials). Subscribe here: https://stat.ethz.ch/mailman/listinfo/bioc-devel

Registration

• support site registration: Unable to connect to support site: Peer certificate cannot be authenticated with given CA certificates: [support.bioconductor.org] SSL certificate problem: certificate has expired