RNA_editing_drugresponse_breastcancer

I. Differential A>I(G) RNA-edited sites (DES) from Breast Cancer cell lines

Data Filters for the identification of differential A>I(G) RNA-edited sites (DES)

Base counts per site with allele AG or TC.

Apply filters for the identification of RNA edited sites (Master_table_CCLE, see code):

a. (Coverage at position >= 10)
b. (Variant == "AG" OR Variant == "TC")
c. (Coverage of alternative allele >= 5)
d. (Frequency of alternative allele >= 0.05)

Merge cell line data into three groups (PARPi, Anthracyclines, Alkylating agents).

DNA:RNA pair for to exclude variant in DNA

Exclude all sites reported in DepMap (using version 19Q3:hg19) (https://depmap.org/portal/download/all/?releasename=DepMap+Public+19Q3)
Downloaded sample_info.csv & CCLE_mutations.csv

a. Load data from DepMapPortal (https://depmap.org/portal/download/all/?releasename=DepMap+Public+22Q2&filename=sample_info.csv)
b. Load data 19Q3:hg19 from DepMapPortal (https://depmap.org/portal/download/all/?releasename=DepMap+Public+19Q3&filename=CCLE_mutations.csv)

PARP inhibitors

DNA data

a. 1,239,235 total
b. 35,520 lineage == "breast" c. 14,397 selected cell lines
d. 2,949 Reference_Allele == "A" | Reference_Allele == "T"
e. 1,450 Reference_Allele == "A" & Strand == "+" | Reference_Allele == "T" & Strand == "-"
f. 1,394 Variant_Type == "SNP"
g. 630 allele == "A/G" & Strand == "+" | allele == "T/C" & Strand == "-"
h. 225 match with 53,043 (position)

RNA data from Master table

53,043 A>I(G) RNA edited sites - 225 variants in DNA = 52,818 RNA-edited sites

Anthracyclines Group:

DNA data

a. 1,239,235 total
b. 35,520 lineage == breast
c. 9,386 selected cell lines
d. 1,703 Reference_Allele == "A" | Reference_Allele == "T"
e. 840 Reference_Allele == "A" & Strand == "+" | Reference_Allele == "T" & Strand == "-"
f. 795 Variant_Type == "SNP"
g. 508 allele == "A/G" & Strand == "+" | allele == "T/C" & Strand == "-"
h. 193 match with 53,043 (position)

RNA data from Master table

54,760 A>I(G) RNA edited sites - 193 variants in DNA = 54,567 RNA-edited sites

Alkylating agents Group:

DNA data

a. 1,239,235 total
b. 35,520 lineage == breast
c. 10,751 selected cell lines
d. 2,119 Reference_Allele == "A" | Reference_Allele == "T"
e. 1,039 Reference_Allele == "A" & Strand == "+" | Reference_Allele == "T" & Strand == "-"
f. 988 Variant_Type == "SNP"
g. 639 allele == "A/G" & Strand == "+" | allele == "T/C" & Strand == "-"
h. 237 match with 53,043 (position)

RNA data from Master table

Apply REDITs test (https://github.com/gxiaolab/REDITs)

From the output, we created this variables for the selection of sites differentially edited

data$EL_R <- data$mle_group_resistance_alpha/(data$mle_group_resistance_alpha + data$mle_group_resistance_beta) 
data$EL_S <- data$mle_group_sensitive_alpha/(data$mle_group_sensitive_alpha + data$mle_group_sensitive_beta) 
data$delta <- data$EL_R - data$EL_S
data$fdr <- p.adjust(data$pvalue, method="BH")
data$class <- ifelse(data$delta >= 0, "Low_sensitivity", "High_sensitivity")
data$fold_change = ((data$EL_R/data$EL_S))
data$fold_change_log = (log(data$fold_change, 2))



data$differential <- ifelse(data$pvalue < 0.01 & data$fdr < 0.1 & data$fold_change_log >= 2.5, 
                            "low_sensitivity",
                            ifelse(data$pvalue < 0.01 & data$fdr < 0.1 & data$fold_change_log <= - 2.5, 
                                   "high_sensitivity", 
                                   "non_differential"))

Characterization of A>I(G) RNA-edited sites

a. Annotate and predict the functional consequences with VEP hg19 from Variant effect Predictor (https://grch37.ensembl.org/Homo_sapiens/Tools/VEP): our results by filtering for one selected consequence per variant allele.
b. Cross-reference with COSMIC GENE CONSENSUS (https://cancer.sanger.ac.uk/census)
c. RepeatMasker version hg19 from (http://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/rmsk.txt.gz)

Summary of sites identified

summary	PARPi	Anthracyclines	Alkylating agents
n group	L= 10 ; H= 8	L= 8 ; H=7	L=10 ; H=9
All A>I(G)	53,043	54,760	52,455
DNA out A/G SNP	52,818	54,567	56,425
diff p-value/FDR/FC	12,347	10,015	13,158
diff L	6,939	5,320	6,047
diff H	5,408	4,695	7,111

II. A>I(G) RNA-edited sites in women with breast cancer

Identification of RNA editing level

Liftover

35,714 DES (no-unique) were convert to hg38 from hg19 in UCSC (https://genome.ucsc.edu/cgi-bin/hgLiftOver)

Machine learning algorithms using A>I(G) RNA edited sites and clinical data