P. aeruginosa 20190720: Cross reference putative SNPs across samples.
atb abelew@gmail.com
2019-07-29
1 Cross reference some putative variants across multiple samples
A quick introduction. I received 5 samples of DNA sequencing data with a question: are there some shared/unique variants compared to the reference genome. As a reminder, these are orn deletion strains of pseudomonas.
I did the following:
- Trim the data aggressively to attempt to make sure there are as few sequencing mishaps as possible left in the data. This includes trimming off the first 10 nucleotides of every read.
- Pass the data to a mapper (bowtie2).
- Pass those results to my own samtools-based variant finder, which is quite conservative.
- Pass those results to the snippy software package.
The agreement is quite nice between the tools, which is encouraging. My tool is perhaps too conservative and suggests that more coverage is needed. Snippy suggests to me that the coverage is sufficient. I think I agree with snippy at this time.
There are a series of images in the igv/ directory which in theory back up these suppositions. I took one picture for each variant observed by snippy. Now I am going to read the results from it and look to see how many of these variants are shared among all samples vs. are unique.
Depending on what these results look like, I may do some venn or other shenanigans.
samples <- c("070919_21", "070919_32", "070919_54",
"070919_65", "070919_76")
csv_files <- paste0("preprocessing/", samples, "/outputs/snippy_paeruginosa_pa14/snps.csv")snp_df <- data.frame()
for (f in 1:length(csv_files)) {
name <- samples[f]
file <- csv_files[f]
tmpdf <- readr::read_csv(file)
tmpdf[["encoded"]] <- glue::glue("{tmpdf[['POS']]}_{tmpdf[['TYPE']]}_{tmpdf[['REF']]}_{tmpdf[['ALT']]}")
tmpdf[["count"]] <- as.numeric(gsub(x=tmpdf[["EVIDENCE"]], pattern="^\\w+:(\\d+) .*$", replacement="\\1"))
tmpdf <- tmpdf[, c("encoded", "count")]
colnames(tmpdf) <- c("encoded", name)
if (f == 1) {
snp_df <- tmpdf
} else {
snp_df <- merge(snp_df, tmpdf, by="encoded", all=TRUE)
}
}## Parsed with column specification:
## cols(
## CHROM = col_character(),
## POS = col_double(),
## TYPE = col_character(),
## REF = col_character(),
## ALT = col_character(),
## EVIDENCE = col_character(),
## FTYPE = col_character(),
## STRAND = col_character(),
## NT_POS = col_character(),
## AA_POS = col_character(),
## EFFECT = col_character(),
## LOCUS_TAG = col_character(),
## GENE = col_character(),
## PRODUCT = col_character()
## )## Warning: 85 parsing failures.
## row col expected actual file
## 1 -- 14 columns 6 columns 'preprocessing/070919_21/outputs/snippy_paeruginosa_pa14/snps.csv'
## 2 -- 14 columns 6 columns 'preprocessing/070919_21/outputs/snippy_paeruginosa_pa14/snps.csv'
## 3 -- 14 columns 6 columns 'preprocessing/070919_21/outputs/snippy_paeruginosa_pa14/snps.csv'
## 4 -- 14 columns 6 columns 'preprocessing/070919_21/outputs/snippy_paeruginosa_pa14/snps.csv'
## 5 -- 14 columns 6 columns 'preprocessing/070919_21/outputs/snippy_paeruginosa_pa14/snps.csv'
## ... ... .......... ......... ..................................................................
## See problems(...) for more details.## Parsed with column specification:
## cols(
## CHROM = col_character(),
## POS = col_double(),
## TYPE = col_character(),
## REF = col_character(),
## ALT = col_character(),
## EVIDENCE = col_character(),
## FTYPE = col_character(),
## STRAND = col_character(),
## NT_POS = col_character(),
## AA_POS = col_character(),
## EFFECT = col_character(),
## LOCUS_TAG = col_character(),
## GENE = col_character(),
## PRODUCT = col_character()
## )## Warning: 81 parsing failures.
## row col expected actual file
## 1 -- 14 columns 6 columns 'preprocessing/070919_32/outputs/snippy_paeruginosa_pa14/snps.csv'
## 2 -- 14 columns 6 columns 'preprocessing/070919_32/outputs/snippy_paeruginosa_pa14/snps.csv'
## 3 -- 14 columns 6 columns 'preprocessing/070919_32/outputs/snippy_paeruginosa_pa14/snps.csv'
## 4 -- 14 columns 6 columns 'preprocessing/070919_32/outputs/snippy_paeruginosa_pa14/snps.csv'
## 5 -- 14 columns 6 columns 'preprocessing/070919_32/outputs/snippy_paeruginosa_pa14/snps.csv'
## ... ... .......... ......... ..................................................................
## See problems(...) for more details.## Parsed with column specification:
## cols(
## CHROM = col_character(),
## POS = col_double(),
## TYPE = col_character(),
## REF = col_character(),
## ALT = col_character(),
## EVIDENCE = col_character(),
## FTYPE = col_character(),
## STRAND = col_character(),
## NT_POS = col_character(),
## AA_POS = col_character(),
## EFFECT = col_character(),
## LOCUS_TAG = col_character(),
## GENE = col_character(),
## PRODUCT = col_character()
## )## Warning: 102 parsing failures.
## row col expected actual file
## 1 -- 14 columns 6 columns 'preprocessing/070919_54/outputs/snippy_paeruginosa_pa14/snps.csv'
## 2 -- 14 columns 6 columns 'preprocessing/070919_54/outputs/snippy_paeruginosa_pa14/snps.csv'
## 3 -- 14 columns 6 columns 'preprocessing/070919_54/outputs/snippy_paeruginosa_pa14/snps.csv'
## 4 -- 14 columns 6 columns 'preprocessing/070919_54/outputs/snippy_paeruginosa_pa14/snps.csv'
## 5 -- 14 columns 6 columns 'preprocessing/070919_54/outputs/snippy_paeruginosa_pa14/snps.csv'
## ... ... .......... ......... ..................................................................
## See problems(...) for more details.## Parsed with column specification:
## cols(
## CHROM = col_character(),
## POS = col_double(),
## TYPE = col_character(),
## REF = col_character(),
## ALT = col_character(),
## EVIDENCE = col_character(),
## FTYPE = col_character(),
## STRAND = col_character(),
## NT_POS = col_character(),
## AA_POS = col_character(),
## EFFECT = col_character(),
## LOCUS_TAG = col_character(),
## GENE = col_character(),
## PRODUCT = col_character()
## )## Warning: 102 parsing failures.
## row col expected actual file
## 1 -- 14 columns 6 columns 'preprocessing/070919_65/outputs/snippy_paeruginosa_pa14/snps.csv'
## 2 -- 14 columns 6 columns 'preprocessing/070919_65/outputs/snippy_paeruginosa_pa14/snps.csv'
## 3 -- 14 columns 6 columns 'preprocessing/070919_65/outputs/snippy_paeruginosa_pa14/snps.csv'
## 4 -- 14 columns 6 columns 'preprocessing/070919_65/outputs/snippy_paeruginosa_pa14/snps.csv'
## 5 -- 14 columns 6 columns 'preprocessing/070919_65/outputs/snippy_paeruginosa_pa14/snps.csv'
## ... ... .......... ......... ..................................................................
## See problems(...) for more details.## Parsed with column specification:
## cols(
## CHROM = col_character(),
## POS = col_double(),
## TYPE = col_character(),
## REF = col_character(),
## ALT = col_character(),
## EVIDENCE = col_character(),
## FTYPE = col_character(),
## STRAND = col_character(),
## NT_POS = col_character(),
## AA_POS = col_character(),
## EFFECT = col_character(),
## LOCUS_TAG = col_character(),
## GENE = col_character(),
## PRODUCT = col_character()
## )## Warning: 84 parsing failures.
## row col expected actual file
## 1 -- 14 columns 6 columns 'preprocessing/070919_76/outputs/snippy_paeruginosa_pa14/snps.csv'
## 2 -- 14 columns 6 columns 'preprocessing/070919_76/outputs/snippy_paeruginosa_pa14/snps.csv'
## 3 -- 14 columns 6 columns 'preprocessing/070919_76/outputs/snippy_paeruginosa_pa14/snps.csv'
## 4 -- 14 columns 6 columns 'preprocessing/070919_76/outputs/snippy_paeruginosa_pa14/snps.csv'
## 5 -- 14 columns 6 columns 'preprocessing/070919_76/outputs/snippy_paeruginosa_pa14/snps.csv'
## ... ... .......... ......... ..................................................................
## See problems(...) for more details.## [1] 146 5## 070919_21 070919_32 070919_54 070919_65
## Min. : 9.0 Min. : 7.0 Min. :10.0 Min. :10.0
## 1st Qu.:12.0 1st Qu.:12.0 1st Qu.:12.0 1st Qu.:12.0
## Median :14.0 Median :14.0 Median :15.0 Median :14.0
## Mean :16.4 Mean :15.6 Mean :17.2 Mean :17.1
## 3rd Qu.:19.0 3rd Qu.:17.0 3rd Qu.:19.0 3rd Qu.:18.0
## Max. :69.0 Max. :46.0 Max. :58.0 Max. :65.0
## NA's :61 NA's :65 NA's :44 NA's :44
## 070919_76
## Min. :10.0
## 1st Qu.:11.0
## Median :13.0
## Mean :15.5
## 3rd Qu.:17.2
## Max. :40.0
## NA's :62write.csv(file="snp_df.csv", x=snp_df)
binary <- snp_df
idx <- binary > 0
binary[idx] <- 1
idx <- is.na(binary)
binary[idx] <- 0
binary_list <- list()
for (e in colnames(binary)) {
idx <- binary[, e] == 1
positions <- rownames(binary)[idx]
binary_list[[e]] <- positions
}
library(Vennerable)
test <- Vennerable::Venn(Sets=binary_list)
Vennerable::plot(test, doWeights=FALSE)
2 Get positions unique to each sample
## [1] "2341567_del_GC_G" "2341608_del_GC_G" "6070084_snp_T_C"## [1] "3373667_snp_G_C" "5103224_mnp_AG_CA" "5791972_del_CA_C"## [1] "2034305_del_TA_T" "2787829_mnp_CT_AG"
## [3] "2885933_snp_G_C" "3387854_snp_A_C"
## [5] "3387865_complex_ATTA_CTTC" "3387881_complex_ATGT_CTGC"
## [7] "3387941_snp_A_C" "4819712_del_GC_G"
## [9] "945882_del_CG_C" "945892_del_GC_G"
## [11] "96350_snp_T_C" "96391_complex_ACA_GCC"
## [13] "96401_snp_G_C" "96406_complex_TCG_GCC"## [1] "2033089_del_AT_A" "2407435_snp_C_G" "2955357_snp_A_G"
## [4] "3051413_snp_T_C" "3051443_snp_T_C" "3374601_snp_A_G"
## [7] "3670384_snp_A_G" "5256362_snp_N_C" "791890_snp_A_G"## [1] "208430_complex_TCCG_GCCC" "5757525_snp_G_C"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