生信编程直播课程优秀学员作业展示1

setwd('E:\\r\\biotrainee_demo\\class1')#修改工作路径t1 <- Sys.time()#把程序运行之前的时间赋值给t1directory = 'CCDS.current.txt'#把文件名赋值给directorydata <- read.table(directory, sep='\t',                   stringsAsFactors=F, header=T)[c(1,10)]#读取数据并提取出第一和第十列get_gene <-function(data_item){  # 该函数用于apply执行  # 输入的数据为仅含原始数据第1列和第10列的dataframe  # 用apply函数执行后输出的数据为每个基因外显子的坐标，  # 一个基因的所有外显子以逗号分隔组成一个string，所有基因的string组成一个vector  # 用apply函数执行后，最后格式为c('111-112, 115-135, 125-138', '254-258',...)  if (!data_item[2] =='-'){    exon_ranges <- data_item[2]    exon_ranges <- substr(exon_ranges, start=2, stop=nchar(exon_ranges)-1)# 去除首尾的中括号符号  }}get_exon <- function(gene){  # 输入的数据为c('111-112, 115-135', '125-138', '254-258,...')  # 把i号染色体上的所有外显子后在一起，并去除完全相同的外显子  # 输出的数据为c('111-112','115-135', '125-138', '254-258', ...)  exon <- unique(strsplit(gene,", ")[[1]])}get_length <- function(exon){  # 输入的数据为lapply(c('111-112','115-135', '125-138', '254-258', ...),fun)  # 输出结果为左右两坐标之差+1即外显子的长度  loc <- strsplit(exon,"-")[[1]]  a <- as.numeric(loc[2])-as.numeric(loc[1]) +1 #每个外显子的碱基数目  a}exon_length = 0exon_length_items = NULLfor (i in unique(data[,1])){  gene_i <- paste(apply(data[which(data[1]==i & data[2] != '-'),], 1, get_gene),collapse=', ')  exon_i <-  get_exon(gene_i)  exon_i_length <- sapply(exon_i, get_length)  exon_length <- exon_length + sum(exon_i_length)  exon_length_items <- c(exon_i_length, exon_length_items)  names(exon_length_items)[1:length(exon_i_length)] <- i}hist(exon_length_items,xlim=c(0,500),breaks = 20000,      main='Distribution of exon length', xlab='exon length')difftime(Sys.time(), t1, units = 'secs')# 计算执行完成后时间与t1的间隔print(paste('all exons length is',exon_length))

setwd('E:\\r\\biotrainee_demo1')t1 <- Sys.time()directory = 'CCDS.current.txt'# 读取数据并提取第1列和第10列data <- read.table(directory, sep='\t',                   stringsAsFactors=F, header=T)[c(1,10)]get_gene <-function(data_item){  # 用apply执行该函数  # 输入的数据为仅含原始数据第1列和第10列的dataframe  # 输出的数据为c('111-112, 115-135, 125-138', '254-258',...)  if (!data_item[2] =='-'){    exon_ranges <- data_item[2]    exon_ranges <- substr(exon_ranges, start=2, stop=nchar(exon_ranges)-1)  }}get_exon <- function(gene){  # 输入的数据为c('111-112, 115-135, 125-138, 254-258,...')  # 输出的数据为c('111-112','115-135', '125-138', '254-258', ...)  exon <- unique(strsplit(gene,", ")[[1]])# 注：strsplit的输出结果为列表}get_length <- function(exon){  # 输入的数据为lapply(c('111-112','115-135', '125-138', '254-258', ...),fun)  # 输出结果为两坐标值和左右两坐标之差  loc <- strsplit(exon,"-")[[1]]  a <- c(as.numeric(loc[1]), as.numeric(loc[2])-as.numeric(loc[1]), as.numeric(loc[2]))  #if (a==0){  #print(loc)  #}  a}exon_length = NULLfor (i in unique(data[,1])){  # paste 函数把i号染色体的所有外显子的坐标合并为一个character对象  # gene_i的格式为'111-112, 115-135, 125-138, 254-258,...'  gene_i <- paste(apply(data[which(data[1]==i & data[2] != '-'),], 1, get_gene),collapse=', ')  # exon_i的格式为c('111-112','115-135', '125-138', '254-258', ...)  exon_i <-  lapply(get_exon(gene_i), get_length)  mat <- matrix(unlist(exon_i), ncol=3, byrow = T)  #mat <- mat[order(mat[,2], decreasing = F),]  #mat <- mat[order(mat[,1], decreasing = F),]  # 使用matrix 是因为vector太长会报错  #R memory management / cannot allocate vector of size n MB  base_loc <- matrix(unique(unlist(apply(mat, 1, function(x) c(x[1]:x[3])))))  exon_length <- c(exon_length , dim(base_loc)[1] * dim(base_loc)[2])}# 耗时长度difftime(Sys.time(), t1, units = 'secs')chrs <- unique(data[,1])barplot(exon_length,names.arg=chrs,xlab='Chromosomes',ylab='length of exons')print(paste('all exons length is',sum(exon_length)))

import pandas as pdimport numpy as npfile = r'E:\r\biotrainee_demo1\CCDS.current.txt'def calculate_exon(file):  data = pd.read_csv(file, sep='\t',\    usecols=[0,9])#data.loc[1:10,:]#  data[0:3]#  data.iloc[1:3]#  data.iloc[3]  all_length = 0  for i in data.iloc[:,0].unique():    # get the data of chrosome i    # iloc[row_vector,col_vect]    # iloc[row_vector]    data_i = data.loc[data.iloc[:,0] == i]    type(data_i)    type(data_i.iloc[:,1])    # remove the '[]' in column2    data_j = data_i.iloc[:,1].apply(lambda x: x[1:-1])    data_p = data_j.apply(lambda x: x.split(', '))    data_g = data_p.apply(lambda x: pd.Series(x))    # 把nan填充为 0-0    data_f = np.array(data_g.fillna('0-0'))    # 去除重复的外显子    data_f = np.unique(data_f.reshape((data_f.shape[0]*data_f.shape[1], 1)))    data_f = pd.DataFrame(data_f)    data_m = data_f.apply(lambda x: \      x.apply(lambda y: (y.split('-')[0])))    data_n = data_f.apply(lambda x: \      x.apply(lambda y: (y.split('-')[-1])))    # pd.to_numeric can only apply to a 1-d array    data_mi = data_m.apply(lambda x: pd.to_numeric(x, downcast='float'))    data_ni = data_n.apply(lambda x: pd.to_numeric(x, downcast='float'))    all_length += (data_ni - data_mi).sum().sum()  return(all_length)length = calculate_exon(file)print(length)