| Type: | Package |
| Title: | Tumor Clones Percentage Estimations |
| Version: | 1.0.1 |
| Author: | Xuan You <youxuan90@gmail.com>, Yichen Cheng <ycheng11@gsu.edu> |
| Maintainer: | Xuan You <youxuan90@gmail.com> |
| Description: | Includes R functions for the estimation of tumor clones percentages for both snp data and (whole) genome sequencing data. See Cheng, Y., Dai, J. Y., Paulson, T. G., Wang, X., Li, X., Reid, B. J., & Kooperberg, C. (2017). Quantification of multiple tumor clones using gene array and sequencing data. The Annals of Applied Statistics, 11(2), 967-991, <doi:10.1214/17-AOAS1026> for more details. |
| License: | GPL-2 | GPL-3 [expanded from: GPL (≥ 2)] |
| Encoding: | UTF-8 |
| LazyLoad: | yes |
| LazyData: | yes |
| Imports: | Rcpp (≥ 0.12.12), PSCBS |
| LinkingTo: | Rcpp, RcppArmadillo |
| RoxygenNote: | 6.0.1 |
| NeedsCompilation: | yes |
| Packaged: | 2018-09-13 00:31:16 UTC; xuan |
| Repository: | CRAN |
| Date/Publication: | 2018-09-13 04:20:02 UTC |
Return mixture estimation of a normal and a tumor Takes BAF, LRR, chr, x, gt, seg_raw
Description
Return mixture estimation of a normal and a tumor Takes BAF, LRR, chr, x, gt, seg_raw
Usage
calc_1d(BAF, LRR, chr, x, GT, seg_raw)
Arguments
BAF |
vector containing B allen frequency (BAF) |
LRR |
vector |
chr |
vector |
x |
vector |
GT |
vector of factors containing genotype |
seg_raw |
dataframe about segmentation |
Value
sol1 |
percentage of tumor for optimal solution 1 |
sol2 |
percentage of tumor 1 for optimal solution 2 |
Return mixture estimation for percentage of normal cells and tumor (1 normal + 1 tumor) with wgs data
Takes baf, lrr, n_baf and nrc
Description
Return mixture estimation for percentage of normal cells and tumor (1 normal + 1 tumor) with wgs data
Takes baf, lrr, n_baf and nrc
Usage
calc_1d_wgs(baf, lrr, n_baf, nrc)
Arguments
baf |
a numeric vector. Each element is the mean adjusted B allele frequency for that segment, calculated as the mean of baf_tumor/baf_normal/2 for that segment |
lrr |
a numeric vector. Each element is the log ratio of the tumor read count and the normal read count for a segment, defined as log(tumorCount/normalCount) |
n_baf |
a numeric vector |
nrc |
a numeric vector. Each element is the normal read count of the segment divided by two |
Value
sol1 |
a numeric number. It provides the estimated percentages of normal from the best solution. The number is the percentage of the estimated normal percentage. |
sol2 |
a numeric number. It provides the estimated percentages of normal from the second best solution. The number is the percentage of the estimated normal percentage. |
Return mixture estimation of a normal and 2 tumors Takes BAF, LRR, chr, x, gt, seg_raw
Description
Return mixture estimation of a normal and 2 tumors Takes BAF, LRR, chr, x, gt, seg_raw
Usage
calc_2d(BAF, LRR, chr, x, GT, seg_raw)
Arguments
BAF |
vector containing B allen frequency (BAF) |
LRR |
vector |
chr |
vector |
x |
vector |
GT |
vectors of factors containing genotype |
seg_raw |
dataframe about segmentation |
Value
sol1 |
a numeric vector of length 2. It provides the estimated percentages of normal and tumor from the best solution. The first number is the percentage of the estimated normal percentage. The second number-1 is the percentage of the estimated tumor 1 percentage |
sol2 |
a numeric vector of length 2. It provides the estimated percentages of normal and tumor from the second best solution. The first number is the percentage of the estimated normal percentage. The second number-1 is the percentage of the estimated tumor 1 percentage |
Return mixture estimation for percentage of normal cells and tumor (1 normal + 2 tumors) with wgs data
Takes baf, lrr, n_baf and nrc
Description
Return mixture estimation for percentage of normal cells and tumor (1 normal + 2 tumors) with wgs data
Takes baf, lrr, n_baf and nrc
Usage
calc_2d_wgs(baf, lrr, n_baf, nrc)
Arguments
baf |
a numeric vector. Each element is the mean adjusted B allele frequency for that segment, calculated as the mean of baf_tumor/baf_normal/2 for that segment |
lrr |
a numeric vector. Each element is the log ratio of the tumor read count and the normal read count for a segment, defined as log(tumorCount/normalCount) |
n_baf |
a numeric vector |
nrc |
a numeric vector. Each element is the normal read count of the segment divided by two |
Value
sol1 |
a numeric vector of length 2. It provides the estimated percentages of normal and tumor from the best solution. The first number is the percentage of the estimated normal percentage. The second number is the percentage of the estimated tumor 1 percentage |
sol2 |
a numeric vector of length 2. It provides the estimated percentages of normal and tumor from the second best solution. The first number is the percentage of the estimated normal percentage. The second number is the percentage of the estimated tumor 1 percentage |
Return mixture estimation of a normal and 3 tumors Takes BAF, LRR, chr, x, gt, seg_raw
Description
Return mixture estimation of a normal and 3 tumors Takes BAF, LRR, chr, x, gt, seg_raw
Usage
calc_3d(BAF, LRR, chr, x, GT, seg_raw)
Arguments
BAF |
vector containing B allen frequency (BAF) |
LRR |
vector |
chr |
vector |
x |
vector |
GT |
vector of factors containing genotype |
seg_raw |
dataframe about segmentation |
Value
sol1 |
percentage of tumor for optimal solution 1 |
sol2 |
percentage of tumor 1 for optimal solution 2 |
return sqrt(n)
Description
return sqrt(n)
Usage
calc_n(n)
calculate baf_1d for 1 normal + 1 tumor case
Description
calculate baf_1d for 1 normal + 1 tumor case
Usage
calcll_1d_baf(lrr, nrc, baf, n_baf, lprior_f_2d, rlprior_f_2d, scale, pscnMax,
MaxCn)
calculate baf for 1 normal + 2 tumors case
Description
calculate baf for 1 normal + 2 tumors case
Usage
calcll_baf(lrr, nrc, baf, n_baf, lprior_f_2d, rlprior_f_2d, scale, pscnMax,
MaxCn)
calculate likelihood for 1 normal + 2 tumors case
Description
calculate likelihood for 1 normal + 2 tumors case
Usage
calcll_cpp(IT_new, B_new, lp, rlp, var_baf, var_tcn, scale, pscnMax, cnMax)
calculate likelihood for 1 normal + 1 tumor case
Description
calculate likelihood for 1 normal + 1 tumor case
Usage
calcll_p1_cpp(IT_new, B_new, lp, rlp, var_baf, var_tcn, scale, pscnMax, cnMax)
calculate likelihood for 1 normal + 3 tumors case
Description
calculate likelihood for 1 normal + 3 tumors case
Usage
calcll_p3(IT_new, B_new, var_baf, var_tcn, scale, pscnMax, cnMax)
combine close segmentation
Description
combine close segmentation
Usage
combine_close_seg(seg_raw, var_baf, data, delta)
It is a function that takes the LRR obtained from SNP array data and returns the estimated tumor and normal proportions. Currently, the function can performs the proportion estimations by assuming the number of tumor clones to be 1 or 2 or 3. The normalization step is not required and the normalization constant will be returned by this function. The function will output two sets of solutions corresponding to the top 2 optimal solutions based on the posterior distribution. You can choose according to your expertise the one that is more reasonable.
Description
It is a function that takes the LRR obtained from SNP array data and returns the estimated tumor and normal proportions. Currently, the function can performs the proportion estimations by assuming the number of tumor clones to be 1 or 2 or 3. The normalization step is not required and the normalization constant will be returned by this function. The function will output two sets of solutions corresponding to the top 2 optimal solutions based on the posterior distribution. You can choose according to your expertise the one that is more reasonable.
Usage
est_mixture(BAF, LRR, chr, x, GT, seg_raw = "NA", num_tumor = 1)
Arguments
BAF |
a numeric vector containing the B Allele Frequency for the sample, corresponding to the location (chr, x). |
LRR |
a numveric vector containing the Log R ratio for the sample, corresponding to the location (chr, x). In practice, the LRR values you include should be the raw LRR output devided by 0.55. |
chr |
a factor vector containing the chromosome. |
x |
a numeric vector containing the location on the chromosome, measured by base pair. |
GT |
a factor vector containing the genotype. Possible values are "AA", "AB", "BB" and NA. |
seg_raw |
Optional. A dataframe containing the segmentaiton results. If not supplied, function |
num_tumor |
1 or 2 or 3, indicating the number of tumor clones. 1 indicates a mixture for a normal and one tumor clone. 2 indicates a mixture for a normal and 2 tumors and so on. Default value is set to be 1. |
Value
sol1_pct |
the estimated percentages for all tumor clones for optimal solution 1. Each value is between 0 and 100. |
sol1_scale |
a scaler that provide the normalization constant for LRR for optimal solution 1. That is 2*2^LRR/scale will be on the same scale as the copy number. |
sol1_cn1 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 1 for the optimal solution. |
sol1_cn2 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 2 for the optimal solution. |
sol1_pscn1 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 1 for the optimal solution. |
sol1_pscn2 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 2 for the optimal solution. |
sol2_pct |
the estimated percentages for all tumor clones for optimal solution 2. Each value is between 0 and 100. |
sol2_scale |
a scaler that provide the normalization constant for LRR for optimal solution 2. That is 2*2^LRR/scale will be on the same scale as the copy number. |
sol2_cn1 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 1 for the second optimal solution. |
sol2_cn2 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 2 for the second optimal solution. |
sol2_pscn1 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 1 for the second optimal solution. |
sol2_pscn2 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 2 for the second optimal solution. |
Examples
##########################################################
##
## short example
##
#########################################################
## first load the data
BAF <- example_data$BAF
LRR <- example_data$LRR ## In practice, the orignal LRR should be devided by 0.55
chr <- example_data$chr
loc <- example_data$x
GT <- example_data$GT
gt = (GT=='BB')*2+(GT=='AB')*1.5+(GT=='AA')-1;gt[gt==(-1)]=NA
## then perform segmentation
gaps = PSCBS::findLargeGaps(x=loc,minLength=5e6,chromosome=chr)
if(!is.null(gaps)) knownSegments = PSCBS::gapsToSegments(gaps)
p <- 0.0001
fit <- PSCBS::segmentByPairedPSCBS(CT=2*2^LRR,betaT=BAF,muN=gt,chrom=chr,
knownSegments=knownSegments,tbn=FALSE,x=loc,seed=1, alphaTCN=p*.9,alphaDH=p*.1)
seg_eg = fit$output
## then perform tumor mixture estimation by assuming 1 tumor clones
out = est_mixture(BAF, LRR, chr, loc, GT, num_tumor = 1, seg_raw = seg_eg)
out$sol1_pct
out$sol1_scale
## References: Quantification of multiple tumor clones using gene array and sequencing data.
## Y Cheng, JY Dai, TG Paulson, X Wang, X Li, BJ Reid, C Kooperberg.
## Annals of Applied Statistics 11 (2), 967-991
## Segmentation-based detection of allelic imbalance and loss-of-heterozygosity
## in cancer cells using whole genome SNP arrays.
## J Staaf, D Lindgren, J Vallon-Christersson, A Isaksson, H Goransson, G Juliusson,
## R Rosenquist, M H, A Borg, and M Ringner
It is a function that takes the count data obtained from whole genome sequencing (WGS) data and returns the estimated tumor and normal proportions. Currently, the function can performs the proportion estimations by assuming the number of tumor clones to be 1 or 2. The normalization step is not required and the normalization constant will be returned by this function. The function will output two sets of solutions corresponding to the top 2 optimal solutions based on the posterior distribution. You can choose according to your expertise the one that is more reasonable.
Description
It is a function that takes the count data obtained from whole genome sequencing (WGS) data and returns the estimated tumor and normal proportions. Currently, the function can performs the proportion estimations by assuming the number of tumor clones to be 1 or 2. The normalization step is not required and the normalization constant will be returned by this function. The function will output two sets of solutions corresponding to the top 2 optimal solutions based on the posterior distribution. You can choose according to your expertise the one that is more reasonable.
Usage
est_mixture_wgs(exp_data, normal_snp, tumor_snp, f_path, num_tumor = 1)
Arguments
exp_data |
a string. It provides the file name of interval. exp_data.intervals should be the name of the interval file. For the format of this file, please see the example section. The file should contain 6 and only 6 columns with each column corresponds to "ID","chrm","start","end","tumorCount" and "normalCount". It is very important to keep the order of the columns the same as listed. |
normal_snp |
a string. It provides the file name of WGS count data for a normal sample or a control sample. |
tumor_snp |
a string. It provides the file name of WGS count data for the tumor sample. |
f_path |
a string. It provides the absolute path of the folder that contains the files above. |
num_tumor |
1 or 2, indicating the number of tumor clones. 1 indicates a mixture for a normal and one tumor clone. 2 indicates a mixture for a normal and 2 tumors and so on. Default value is set to be 1. |
Value
sol1_pct |
the estimated percentages for all tumor clones for optimal solution 1. Each value is between 0 and 100. |
sol1_scale |
sol1_scale: a scaler that provide the normalization constant for LRR for optimal solution 1. That is 2*tumor_count/normal_count will be on the same scale as the copy number. |
sol1_cn1 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 1 for the optimal solution. |
sol1_cn2 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 2 for the optimal solution. |
sol1_pscn1 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 1 for the optimal solution. |
sol1_pscn2 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 2 for the optimal solution. |
sol2_pct |
the estimated percentages for all tumor clones for optimal solution 2. Each value is between 0 and 100. |
sol2_scale |
sol1_scale: a scaler that provide the normalization constant for LRR for optimal solution 2. That is 2*tumor_count/normal_count will be on the same scale as the copy number. |
sol2_cn1 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 1 for the second optimal solution. |
sol2_cn2 |
a vector of length S, where S is the number of segments. It is the estimated copy number for tumor 2 for the second optimal solution. |
sol2_pscn1 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 1 for the second optimal solution. |
sol2_pscn2 |
a vector of length S, where S is the number of segments. It is the estimated parent specifit copy number for tumor 2 for the second optimal solution. |
Examples
exp_data = "data_exp_eg" ## exp_data.intervals should be the file name of the segments.
## For the format of the input files, you can use the example code below.
normal_snp = "snp_norm_eg" ## snp_norm_eg.txt should be the count file name for the normal sample.
tumor_snp = "snp_tum_eg" ## snp_tum_eg.txt should be the count file name for the tumor sample.
f_path = system.file("extdata",package="EstMix")
## f_path should be the absolute path of folder that contains the txt and interval files.
out_wgs = est_mixture_wgs(exp_data, normal_snp, tumor_snp,f_path,num_tumor = 1)
out_wgs$sol1_pct
out_wgs$sol1_scale
## for the format of the input files, please see the following code
data_exp_path = file.path(f_path, paste("/", exp_data, ".intervals", sep=""))
snp_norm_path = file.path(f_path, paste("/",normal_snp, ".txt", sep=""))
snp_tumor_path = file.path(f_path, paste("/",tumor_snp, ".txt", sep=""))
data_exp = read.table(data_exp_path);
colnames(data_exp) = c("ID","chrm","start","end","tumorCount","normalCount")
snp_norm = read.table(snp_norm_path)
snp_tum = read.table(snp_tumor_path)
## References: Quantification of multiple tumor clones using gene array and sequencing data.
## Y Cheng, JY Dai, TG Paulson, X Wang, X Li, BJ Reid, C Kooperberg.
## Annals of Applied Statistics 11 (2), 967-991
ExampleData
Description
ExampleData
Usage
example_data
Format
A data frame with 5 variables: BAF, chr, GT, LRR and x, with chromosome number = 22
f_baf
Description
f_baf
Usage
f_baf(x, beta, var = 8e-04)
calculate segmentation
Description
calculate segmentation
Usage
get_segmentation(BAF, LRR, chr, x, GT)
get_var_tcn_baf
Description
get_var_tcn_baf
Usage
get_var_tcn_baf(LRR_raw, BAF_raw, gt, sl = 1000)
trancate data and take mean
Description
trancate data and take mean
Usage
mean2(x)
Preprocessing data
Description
Takes exp_data, normal_snp, and tumor_snp
Usage
preprocessing(exp_data, normal_snp, tumor_snp, f_path)
Arguments
exp_data |
a string, file name of xxx |
normal_snp |
a string, file name of xxx |
tumor_snp |
a string, file name of xxx |
f_path |
a string, file path of the files above |
Value
df |
a dataframe containing lrr, nrc, baf and n_baf |
Segmentation
Description
Segmentation
Usage
seg_eg
Format
A data frame with segmentation info, with chromosome = 22
select scale for 1 normal + 1 tumor case
Description
select scale for 1 normal + 1 tumor case
Usage
sel_scale_1d(cnMax = 6, pscnMax = 6, ngrid = 100, nslaves = 50, B_new,
IT_new, temp)
select scale for 1 normal + 1 tumor case for wgs data
Description
select scale for 1 normal + 1 tumor case for wgs data
Usage
sel_scale_1d_wgs(cnMax = 6, pscnMax = 6, ngrid = 100, nslaves = 50,
temp)
select scale for 1 normal + 2 tumors case
Description
select scale for 1 normal + 2 tumors case
Usage
sel_scale_2d(cnMax = 6, pscnMax = 6, ngrid = 100, nslaves = 50, B_new,
IT_new, temp)
select scale for 1 normal + 2 tumors case for wgs data
Description
select scale for 1 normal + 2 tumors case for wgs data
Usage
sel_scale_2d_wgs(cnMax = 6, pscnMax = 6, ngrid = 100, nslaves = 50,
temp)
wgsData
Description
wgsData
Usage
wgs_eg
Format
A data frame with 4 variables: baf, lrr, n_baf and nrc