I have a data frame in R with three columns: Gene.ID, source, and value. I need to filter the rows based on multiple conditions, but I'm having trouble achieving the desired result. Here's a sample of my data:
My goal is to:

我在R中有一个数据框，它有三列：Gene.ID、SOURCE和VALUE。我需要基于多个条件来筛选行，但我很难获得所需的结果。以下是我的数据样本：我的目标是：

Keep rows with the same Gene.ID and source.
For rows with the same Gene.ID but different source, I want to keep them only if the value is different from the previous row.
I've tried various approaches using dplyr and custom loops, but I haven't been able to achieve the desired filtering logic.

保留具有相同Gene.ID和源的行。对于具有相同Gene.ID但来源不同的行，我希望仅当值与前一行不同时才保留它们。我尝试了使用dplyr和定制循环的各种方法，但未能实现所需的过滤逻辑。

Can someone provide a solution or suggest an efficient way to filter this data frame based on these conditions?

是否有人可以提供解决方案或建议一种有效的方法来根据这些条件过滤此数据框？

Thank you for your assistance!

感谢您的帮助！

df <- data.frame(
  Gene.ID = c(
    "NZ_JAHWGH010000001.1_15",
    "NZ_JAHWGH010000001.1_17",
    "NZ_JAHWGH010000001.1_68",
    "NZ_JAHWGH010000001.1_7"
  ),
  HMMER = c(
    "SLH",
    "GT2",
    "GT2",
    "GH13+CBM41+CBM41+GH13"
  ),
  dbCAN_sub = c(
    "",
    "GT2",
    "GT2",
    "CBM41+GH13+CBM41+CBM41+CBM48+GH13"
  ),
  DIAMOND = c(
    "",
    "",
    "GT2",
    "CBM41+CBM48+GH13+GH13+GH11"
  ),
  stringsAsFactors = FALSE
)

my desired output will be as followes

我想要的输出如下

df_output <- data.frame(
  Gene.ID = c(
    "NZ_JAHWGH010000001.1_15",
    "NZ_JAHWGH010000001.1_17",
    "NZ_JAHWGH010000001.1_68",
    "NZ_JAHWGH010000001.1_7",
    "NZ_JAHWGH010000001.1_7",
    "NZ_JAHWGH010000001.1_7",
    "NZ_JAHWGH010000001.1_7",
    "NZ_JAHWGH010000001.1_7",
    "NZ_JAHWGH010000001.1_7",
    "NZ_JAHWGH010000001.1_7"
  ),
  combined = c(
    "SLH",
    "GT2",
    "GT2",
    "CBM41",
    "GH13",
    "CBM41",
    "CBM41",
    "CBM48"
    "GH13",
    "GH11"
  ),
  stringsAsFactors = FALSE
)

I tried with this command But I didnt get, desired output

我尝试使用此命令，但没有得到所需的输出

df_output <- df %>%
  separate_rows(., sep = "\\+") %>%
  gather(key = "source", value = " ", -Gene.ID) %>%
  filter(combined != "") %>%
  distinct(Gene.ID, combined)

A working version following your approach replacing some of the tidyr functions with more recent versions:

遵循您的方法的工作版本用更新的版本替换了一些tidyr函数：

library(dplyr)
library(tidyr)

df %>%
  tidyr::separate_longer_delim(HMMER, delim = "+") %>%
  tidyr::separate_longer_delim(dbCAN_sub, delim = "+") %>%
  tidyr::separate_longer_delim(DIAMOND, delim = "+")%>%
  tidyr::pivot_longer(-Gene.ID, values_to = "combined") %>%
  dplyr::select(Gene.ID, combined) %>%
  dplyr::filter(combined != "") %>%
  dplyr::distinct()


# A tibble: 7 x 2
  Gene.ID                 combined
  <chr>                   <chr>   
1 NZ_JAHWGH010000001.1_15 SLH     
2 NZ_JAHWGH010000001.1_17 GT2     
3 NZ_JAHWGH010000001.1_68 GT2     
4 NZ_JAHWGH010000001.1_7  GH13    
5 NZ_JAHWGH010000001.1_7  CBM41   
6 NZ_JAHWGH010000001.1_7  CBM48   
7 NZ_JAHWGH010000001.1_7  GH11 

Note that this might create very long intermediate dataframes if run with a larger initial dataframe, as the separate_longer steps create many rows with redundant content which are only dropped again at the end (in this example 4 -> 359 -> 7 rows). There is probably a more efficient way to do this for large dataset.

请注意，如果使用较大的初始数据帧运行，这可能会创建非常长的中间数据帧，因为Separate_Long步骤创建了许多具有冗余内容的行，这些内容只会在结束时再次删除(在本例中为4->359->7行)。对于大型数据集，可能有一种更有效的方法来实现这一点。