Improving gene function predictions using independent transcriptional components - Raw Figure Data

>*****_medians_file

gene_set_name : Name of gene set

size : Number of genes with prediction scores in that gene set

gene_set_db : Name of host gene set collection

Method : Method used to calculate prediction scores

Subset : Median was calculated using member genes (n = size) or genes that are never members in that gene set collection.

median_prediction_score: median prediction score (for gene sets with less than 10 or more than 500 genes = NA)

>******_multifunctonality_file

gene_set_name : Name of gene set

multifunctionality_score_correlation : Distance correlation between member prediction scores and multifucntionality score calculated using host gene set collection (Empty value = gene set size was outside 10->500 range)

gene_set_db : Name of host gene set collection

method : Method used to calculate prediction scores

>*****_old_version_comparison_file

gene_set_name : Name of gene set

size : Number of genes that were added between v3.0 and v6.2

gene_set_db : Name of host gene set collection

type : version of gene set used to calculate prediction scores of these genes

method : Method used to calculate prediction scores

Median prediction score : median prediction score of subset of genes added in between v3.0 and v6.2

>unbiased_clustering

File with probe entrez_id to cluster mapping

GROUP : cluster number

LABEL : Affymetrix probe

ENTREZID : Corresponding entrez number

SYMBOL : Corresponding symbol

uncharacterized : 0 or 1 if its an orf or LOC gene

>cluster_predictability

File with cluster metrics

GROUP : cluster number

size : size of cluster

median_max_prediction : Maximum prediction score for each gene across 16 collections, cluster median.

density : density metric

median_multifunctionality : Median disctance correlation association to multifunctionality calculated using all genesets from all collections

ORF or LOC : amount of orf or LOC genes in cluster

>durocher_comparison_figure_data

Comparison data for Olivieri et al hit genes

entrez_id : entrez number

variable : Comparison is GO_DNA_REPAIR to GO_DNA_REPAIR

pca_prediction_scores : PCA based prediction score

i.variable : Comparison is GO_DNA_REPAIR to GO_DNA_REPAIR

ica_prediction_scores : ICA based prediction score

known_link_to_DDR : 0 or 1 if it was called as known link in Olivieri et al

>57_lps_network_all_genes

ICA-TC based prediction scores for Immunological Signatures gene set collection, of the subset of genes identified in a CRISPR-Cas9 screen that have a high co-functionality.

entrez_number: Entrez ID of gene

gene_name: HGNC Gene Symbol

gene_set: Immunological Signatures gene set

value: Z-score of gene for gene set

>HALLMARK_ICA_ZTpvalues_CORFsandLOCS_wardClustered

Matrix containing ICA-TC based prediction scores for all Corf and LOC genes, hierarchically clustered using ward's method and 1-cor(dist) as distance function.

Columns correspond to hallmark gene sets, rows to genes.

>HALLMARK_ICA_ZTpvalues_CORFsandLOCS_wardClustered_cutoff_0.8

Cluster membership of the 835 Corf and LOC genes at a dendrogram cutoff height of 0.8

GROUP: Cluster number

LABEL: HGNC Gene Symbol

>HALLMARK_ICAvPCA_ZTpvalues_CORFsandLOCS

Comparison between ICA-TC based and PCA-TC based prediction scores of Hallmark gene sets for all Corf and LOC genes.

gene: HGNC Gene Symbol

variable: gene set

i.value: ICA-TC based prediction score

value: ICA-TC based prediction score

category: logical; 1 if i.value>value, 0 if value>i.value

>ICAvPCA_GO_negviralregulation.txt

Comparison between ICA-TC based and PCA-TC based prediction scores of Hallmark gene sets for all Corf and LOC genes.

entrez_id: Entrez ID of gene

variable: gene set

i.value: ICA-TC based prediction score

value: ICA-TC based prediction score

known_link: one of three strings: "yes" if gene is a member of the gene set; "no" if gene is not a member of the gene set, "screen" if the gene is one of the hits of the investigated CRISPR-cas screen

>ICAvPCA_KEGG_Lysosome

Comparison between ICA-TC based and PCA-TC based prediction scores of Hallmark gene sets for all Corf and LOC genes.

entrez_id: Entrez ID of gene

variable: gene set

i.value: ICA-TC based prediction score

value: ICA-TC based prediction score

known_link: one of three strings: "yes" if gene is a member of the gene set; "no" if gene is not a member of the gene set, "screen" if the gene is one of the hits of the investigated CRISPR-cas screen

>Mouse_v_Human_barcode_spearman_correlations.txt

Spearman correlations of mouse gene barcodes with ortholog human gene barcodes for each of the 16 gene set collections.

mouse_gene: Mouse Entrez ID

assoc_human: Entrez ID of corresponding human ortholog

spearman_r: Spearman correlation coefficient

collection: number corresponding to gene set collection

name: gene set collection name