Basic workflow

About script of generating report of Basic workflow

Initialize object

Set SIRIUS project path and its version to initialize mcnebula object.

## The `path` is where your SIRIUS project saved.
path <- "."
mcn <- mcnebula()
mcn <- initialize_mcnebula(mcn, "sirius.v5", path)
## "pos" or "neg"
ion_mode(mcn) <- "pos"

Filter candidates

Filter molecular formula and chemical structure

According to the top chemical structure to build specific_candidate of reference. Learn more

mcn <- filter_structure(mcn)
mcn <- create_reference(mcn)
mcn <- filter_formula(mcn, by_reference = T)

  • Tips:

    1. Use dplyr::* in filter_structure to customize the filtering, such as: filter_structure(mcn, dplyr::filter, tani.score >= .5). This is similar in filter_formula or filter_ppcp.

  • Tips (after above):

    1. Use latest(mcn, subscript = ".f3_fingerid") to get the filtered candidates of the chemical structure (Top candidate for each feature)
    2. Use latest(mcn, "project_dataset", subscript = ".f3_fingerid") to get all candidates of all features
    3. Use reference(mcn)$specific_candidate to get the selected specific_candidate
    4. Use latest(mcn, subscript = ".f2_formula") to get the filtered candidates of the molecular formula (Top candidate for each feature)
    5. Use latest(mcn, "project_dataset", subscript = ".f2_formula") to get all candidates of all features

Filter chemical classification

Learn more about chemical classification.

mcn <- create_stardust_classes(mcn)
mcn <- create_features_annotation(mcn)
mcn <- cross_filter_stardust(
  mcn, max_ratio = .05,
  cutoff = .4, identical_factor = .6
)
  • Tips (after above):
    1. Use stardust_classes(mcn) to get the data about ‘stardust_classes’
    2. Use features_annotation(mcn) to get the data about …
    3. Use cross_filter_stardust() to get the default parameters about this methods.

(Option) Manually check the filtered chemical classes

Get the kept classes or the filtered out classes.

## The kept classes
classes <- unique(stardust_classes(mcn)$class.name)
## The filtered out classes
table.filtered.classes <- backtrack_stardust(mcn)

(Option) Manually filter the chemical classes

Manually filter some repetitive classes or sub-structural classes. By means of Regex matching, we can obtain a number of recurring name of chemical classes that would contain manay identical compounds as their sub-structure.

## Regex match
pattern <- c("stero", "fatty acid", "pyr", "hydroxy", "^orga")
dis <- unlist(lapply(pattern, grep, x = classes, ignore.case = T))
dis <- classes[dis][-1]
## Remove these classes in `stardust_classes(mcn)`
mcn <- backtrack_stardust(mcn, dis, remove = T)
  • Tips:
    1. Use backtrack_stardust(mcn, dis) without remove to recover these classes.

Create Nebula-Index

After following, whether it is all filtered by the algorithm provided by MCnebula2’s function or custom filtered for some chemical classes, we now have a data called ’nebula_index’. This data records a number of chemical classes and the ‘features’ attributed to them. The subsequent analysis process or visualization will be based on it.

mcn <- create_nebula_index(mcn)
  • Tips (after above):
    1. Use nebula_index(mcn) to get the data …

Create Nebulae

Compute spectral similarity

See help(compute_spectral_similarity) about the algorithm.

mcn <- compute_spectral_similarity(mcn)

Create Parent-Nebula

Create network data for Parent-Nebula.

mcn <- create_parent_nebula(mcn)
mcn <- create_parent_layout(mcn)
  • Tips (after above):
    1. Use function of igraph::write_graph to output .graphml (e.g., use for Cytoscape), such as: igraph::write_graph(igraph(parent_nebula(mcn)), "test.graphml", "graphml")
    2. Use parent_nebula(mcn) to get the data …
    3. Use igraph(parent_nebula(mcn)) to get the data …
    4. …

Create Child-Nebulae

Create network data for Child-Nebulae.

mcn <- create_child_nebulae(mcn)
mcn <- create_child_layouts(mcn)
  • Tips (after above):
    1. See help(create_child_nebulae) for example about how to export ‘.graphml’ of all Child-Nebulae.

Activate the Nebulae for visualization

This would create ‘ggset’ about functions and parameters for ggplot2 to visualize the Nebulae.

mcn <- activate_nebulae(mcn)
  • Tips (after above):
    1. Use function of get_ggset to get the ‘ggset’ of Nebulae. Such as: get_ggset(mcn, 'parent', modify_set_labs).
    2. Use method of call_command to visualize ‘ggset’, such as: call_command(get_ggset(mcn, 'parent', modify_set_labs)).

Visualize Nebulae

Visualize Parent-Nebula 

visualize(mcn, "parent")

Visualize Child-Nebulae 

## Visualize the specific item (number or classes name)
visualize(mcn, 1)
## Visualize all into one plot
visualize_all(mcn)

(Option) Workflow Mode

Use the following command to get the code that outputs the full report containing the above analysis. You can modify the parameters on the basis of the output code to fit your data or your needs.

workflow(mode = "print")
  • Tips: Using sink() to catch the text output in the command lines, such as:
file <- tempfile(fileext = ".R")
sink(file)
workflow(mode = "print")
sink()
← Usage
Metabolic workflow →