SFARI Genes and where to find them; classification modelling to identify genes associated with Autism Spectrum Disorder from RNA-seq data

Code repository for SFARI Genes and where to find them; classification modelling to identify genes associated with Autism Spectrum Disorder from RNA-seq data

Notes about this repository

All code is in R. The drake package is used to manage the workflow of the project, but the code can also be executed as a regular R script:

• The script run.R runs the project as a regular R script and saves the output in the Results folder

• The script make.R runs the project using drake and saves the output in the .drake folder, which can be accessed by name using drake's loadd() function

Note on using drake: Drake provides a lot of useful features but it has two drawbacks in this project:

• Running the project using run.R is much faster than with make.R in computers with multiple cores because some packages use the parallel package underneath, which doesn't work well with clustermq, the package drake uses to distribute the work

• The Enrichment Analysis of the top modules is only available when running the code using run.R, because of compatibility issues between the package clusterProfiler and drake

Running the code

1. Clone this repository

2. Download InputData from doi.org/10.7488/ds/2980

3. Execute run.R or make.R depending on whether you want your workflow to be run drake or not

InputData

• genes_GO_annotations: Gene Ontology annotations for each gene

• krishnan_probability_score.xlsx: Krishnan's ASD probabilty score downloaded from asd.princeton.edu

• NCBI_gene2ensembl_20_02_07gz: NCBI's mapping between genes symbols and ensembl IDs

• NCBI_gene_info_20_02_07_.gz: Functional annotations of the genes

• RNAseq_ASD_datExpr.csv: Gene expression matrix. Downloaded from mgandal's github repository

• RNAseq_ASD_datMeta.csv: Metadata of the samples from the gene expression matrix. Downloaded from mgandal's github repository

• sanders_TADA_score.xlsx Sanders TADA score downloaded from [He et al., 2013)[https://doi.org/10.1371/journal.pgen.1003671]

• SFARI_genes_01-03-2020.csv: SFARI Gene scores using new scoring system

• SFARI_genes_08-29-2019.csv: SFARI Gene scores using old scoring system

Output

Preprocessed Input Data

• new_SFARI_dataset: Dataframe with information about SFARI genes with the new annotation criteria (scores 1 to 3)

• old_SFARI_dataset: Dataframe with information about SFARI genes with the original annotation criteria (scores 1 to 6)

• NCBI_dataset: Dataframe with gene biotype annotation obtained from NCBI

• GO_neuronal_dataset: Dataframe with gene annotation indicating if they have some neuronal-related function in the Gene Ontology

• Gandal_dataset: RData object containing the preprocessed and normalised gene expression data

WGCNA

• modules_dataset: Dataframe indicating the module each of the genes belong to

• top_modules_by_Diagnosis: Dataframe indicating the modules with the highest relation to Diagnosis as well as their correlation value

• top_modules_by_SFARI: Dataframe indicating the modules with the highest enrichment in SFARI Genes as well as their enrichment and adjusted p-value

• top_modules_enrichment: (not included in the drake workflow) Named list with the Enrichment results for all the modules with a strong correlation to Diagnosis or enriched in SFARI Genes

Classification Model

• classification_dataset: Dataframe with the input data used for the classification models

• biased_classification_model: Named list with the information from the biased classification model, including the predictions for each gene and the coefficients and performance metrics of the model

• unbiased_classification_model: Named list with the information from the unbiased classification model. The list includes the same elements as biased_classification_model