bimib-disco / tronco

Repository of the TRanslational ONCOlogy library, which includes various algorithms (such as CAPRESE and CAPRI) and the Pipeline for Cancer Inference (PICNIC).

TRONCO (TRanslational ONCOlogy)

Branch	Status
master
development

The TRONCO (TRanslational ONCOlogy) R package collects algorithms to infer progression models via the approach of Suppes-Bayes Causal Network, both from an ensemble of tumors (cross-sectional samples) and within an individual patient (multi-region or single-cell samples).

The package provides parallel implementation of algorithms that process binary matrices where each row represents a tumor sample and each column a single-nucleotide or a structural variant driving the progression; a 0/1 value models the absence/presence of that alteration in the sample.

The tool can import data from plain, MAF or GISTIC format files, and can fetch it from the cBioPortal for cancer genomics. Functions for data manipulation and visualization are provided, as well as functions to import/export such data to other bioinformatics tools for, e.g, clustering or detection of mutually exclusive alterations. Inferred models can be visualized and tested for their confidence via bootstrap and cross-validation.

TRONCO is used for the implementation of the Pipeline for Cancer Inference (PICNIC).

tronco's Issues

Export - import NBS

Nella export NBS:

• creare anche lo script .m che esegue NBS direttamente dalla export verso NBS assegnando i default per i path dei file usati da NBS
• nello script permettere di selezionare per quali valori di K eseguire il clustering
• aggiungere allo script m il salvataggio di un .mat che contiene una struct matlab con una dataset per ogni valore di K. il dataset deve contenere una colonna con la label ottenuta dal cluster e i nomi delle righe devono essere i sample_id dei pazienti

Nella import NBS:

• avere una funzione che prende in input .mat generato da NBS e crea un dataframe R con una colonna per ogni valore di K, i valori di K nelle righe e come rownales i sample_id dei pazienti

add option to remove node without connection

In tronco.plot add an option to remove node without connection

tronco.plot & caprese

Change the legend syntax to support caprese (remove pv)

consolidate.data

Remove duplicated loops

Import TCGA dataset

Import somatic mutation file (XLSX format) from https://tcga-data.nci.nih.gov/docs/publications/coadread_2012/

empty field must be checked

Columns:

• Hugo_Symbol
• Entrez_Gene_Id ( https://www.biostars.org/p/69647/ )
• Variant_classification
• Tumor_Sample_Barcode

From CRC_clinical_sheet import patient, stage. (NBS needs survival data)

In crc_gistic.txt colums are patients (not samples), short name is used.

oncoprint + annotate.consolidate.events = T

oncoprint(pt, annotate.consolidate.events = T)

Error in genes.annotation[unlist(lapply(invalid$indistinguishable, function(z) { :
incorrect number of subscripts on matrix
In addition: Warning message:
In genes.annotation$consolidate = "none" : Coercing LHS to a list

tmp = consolidate.data(pt, T)

Indistinguishable events:
type event
G22 "variant" "chr19_52888774"
G23 "variant" "chr19_52888833"
Total number of events for these genes: 2

as.events improvements

In as.events add parameters to select genes or types

Plot of prima facie reconstruction

When plotting the prima facie topology, the prima facie only edges should be shown in red

Loop removal

Remove loops on lifted adjacency matrix

Hypo.add

hypotheses.add('1', OR(x), y )
hypotheses.add('2', OR(x), z )

Fails, while it should not - add z as y as possible effects.

tronco.plot + samples annotatation

Not working

pathways @ tronco.plot --> samples.annotation

parameter renaming -- check consistency of naming for all other parameters across semantically related functions such as tronco.plot and oncoprint

Import MAF color

Problem on color palette of import MAF (Accent assigned)

Check estimation in capri

Check estimation in capri

Visualization

Check bug in tronco.plot

Fix in delete.gene

In delete.gene specifically (but we need to check in all the other utility scripts) when removing a gene, it should be removed from the hypotheses too.

Bug fix in overall confidence by bootstrap

Fix the estimation of the overall confidence by bootstrap both with CAPRESE and CAPRI

tronco.plot new legend

remove OR XOR AND prefix from node and improve legend with description of formulas

coditional importation in import.MAF()

change none in no reg

in edmons and prim

handle indistinguishable events

We need to decide whether to merge them or not.

Functions association

To do list for the TRONCO / Cytron magic program!

Check	Function	TRONCO name	Description	Notes
	Import query TCGA		Import data from TCGA	@caravagn from the website?
	Import query CBIO	cbio.query	Import data from CBIO
	Import MAF			maf simple lib
	Import VCF			vcf example
	Import GISTIC	import.gistic
	Import TXT
	Viz table
	Viz oncoprint	oncoprint		rename -> tronco.print ?
	Viz report	show
	Manual Selection	events.selection		samples, genes, events
	Frequency Selection	events.selection		samples, genes, events
	Name I/O Selection	events.selection		samples, genes, events
	Rename	rename.type rename.gene		genes, types
	Delete	delete.type delete.gene		type,genes
	Add			samples, events (bind)
	Merge	union.types -> merge.types		type, samples, genes, events, type=1/type=2
	Bind			samples (sbind), events (ebind)

as.selective.advantage.relations does nothing with caprese models

As it says: using as.selective.advantage.relations on a caprese model will return NULL, even when we can see there are terms for pr, tp and hg (using, say, as.confidence(themodel, conf = c("pr", "tp", "hg"))).

But the help does not indicate that as.selective.advantage will not work with caprese models.

Hypo.add is.compliant

Update $annotations and $types in the input object -- which should be made available at any time by using as.dataset(...). The idea is that I want to plot lifted matrices as well.

is.compliant(x)
x.h = hypothesis.add(x, ...)
oncoprint(as.dataset(x.h))

enforce numeric

Add a call to enforce.numeric in tronco.capri and in tronco.caprese

No edge in adjacency matrix! Nothing to show here.

I'm using TRONCO for progression model on one of our project involving hematological cancer. We have very low mutational load to begin with. Everything seems to work , but during tronco.plot(model), I get

NULL
Warning message:
In tronco.plot(apl.model) :
  No edge in adjacency matrix! Nothing to show here.

Is this because of low mutation numbers ? Any parameters I need to tune in ?

new tronco.plot without hypotheses

Fix bugs in the new version of the tronco.plot when no hypothesis is provided

Vignette

Create a new vignette.

• part 1: data manipulation (CRC)
• part 2: reconstruction (aCML)
• description

oncoprint / show.patterns = T

Check layout, improve it.

Create test dataset

create a random compliant dataset for test purpose

Refactoring

Make the capri output compliant with is.compliant

Inconsistent function call

In import.MAF, at line 110 there is a call to import.genotypes with parameter default.variant = "Mutation", however no such parameter exists in import.genotypes definition.
Same problem is raised in import.GISTIC.
Changing default.variant for event.type in the function call all seems to work so possibly this is the solution.

Change column event (dataset) in atom

all function must be checked

doc - roxygen2 style

• as.functions.R
• caprese.algorithm.R
• caprese.bootstrap.R
• caprese.estimation.R
• capri.algorithm.R
• capri.bootstrap.R
• capri.estimation.R
• capri.hypotheses.R
• correctness.R
• data.R
• editing.functions.R
• external.R
• loading.R
• selection.R
• tronco.R
• visualization.R
• which.functions.R
• zzz.R

as.kfold.posterr returns a data frame with all columns as characters

I do not think that is intended as the columns for MEAN.POSTERR and SD.POSTERR are also of type character, not numeric. In fact, that this is happening can be seen also in the vignette, p. 37. This is inconvenient, and inconsistent with prederr.

Example:

data(test_model_kfold)
s1 <- as.kfold.posterr(test_model_kfold)
summary(s1[[1]])
summary(s1[[2]])

intersect

Fare l'interserct tenendo conto di 2 cose:

• pazienti, presenti sia in cnv sia in mut
• eventi, presente sia in cnv sia in mut se presenti sono stati analizzati, possono anche essere a 0

parametric bootstrap problem

edit foreach statement (num == 1 in parallel env is not a good idea)

env are evil

Remove global env variables in hypothesis functions.

Add print during computation

In function capri.fit and caprese.fit add print with information for the user

Parallelize bootstrap

Parallelize score computations by bootstrap

Atomic to hypothesis

pseudo presudo presudo code:

In hypotheses expansion:

foreach hypo
    if atomic nodes to hypo
        transpose hypo adj matrix
        add prefix to logic nodes (like `UP_AND...`,...)
        initial node <- which column sum == 0
        merge graph on adj matrix and  hypo t adj matrix
        reconnet edges to initial node (from adj matrix search for atomic node to hypo)

In tronco plot:

create layout for UP logic nodes

draw hidden "AND"

see the title

problem on capri.bootstrap

@danro9685 please fix this problem:

check dataset hlist hypotheses in capri.bootstrap

tronco.plot warning when an invalid model is selected

Add a warning where all the available models are enumerated when calling tronco.plot with an invalid model argument.

tronco.plot + samples annotatation

check

remove do.estimation

Remove if (do.estimation) { ...

Error: Possible regularization are bic or aic

When running any algorithm with a regularization, the possible ragularizator loglik, aic and bic.
Here loglik is missing.

remove oncoprint return

And add gtable = FALSE

Manual to be finalized

Remove not exported functions and manual to be finalized.

tronco.plot

mettere ellypsis verso oncoprint?