比较R中的两个数据集

Question

I've searched around here and google for hours and cannot find a solution to my problem.

I have two datasets containing genes. One dataset is my dataset (snap) and I need to see if these genes are in the second bigger data set (catalog). I want the second column in snap (proxy) and the 21st column from catalog. This is what my datasets look like;

> head(snap)
        SNP     Proxy Distance RSquared DPrime
1 rs4246511 rs7540233     4541    0.874      1
2 rs4246511 rs4970634    15768    0.874      1
3 rs4246511 rs4532801    18960    0.874      1
4 rs4246511 rs9438982    22242    0.874      1
5 rs4246511 rs9438979    25034    0.874      1
6 rs4246511 rs4414011    25868    0.874      1 

head(catalog)
        SNPS MERGED SNP_ID_CURRENT    CONTEXT INTERGENIC
1  rs7079041      0        7079041     intron          0
2  rs7244261      0        7244261 intergenic          1
3 rs10448044      0       10448044 intergenic          1
4  rs2610025      0        2610025 intergenic          1
5  rs1472147      0        1472147     intron          0
6  rs2648708      0        2648708     intron          0

*This is a small part of the datasets

To make it more complicated, I'd also like to be able to pull the whole row of data from both data sets.

For the first part of my question I have tried using comparison (which I found from another similar question here). I decided to extract the columns I needed to simplify things (proxy is my column from snap and catalogsnps is the column from catalog);

    comparison <- compare(proxy, catalogsnps, allowAll=TRUE)
comparison$tM

difference <- data.frame(lapply(2:ncol(proxy),function(i)setdiff(cacheGenericsMetaData[,i],comparison$tM[,i])))
colnames(difference) <- colnames(proxy)
write.table(difference, file="difference.csv", sep=";", dec=".")

However with this syntax my output is simply a list of all my SNPs from snap.

Output

1054  6267
1055  6273
1056  6297
1057  6297
1058  6314
1059  6331
1060  6340
1061  6345
1062  6346
1063  6350
1064  6364
1065  6412
1066  6417
1067  6417
1068  6430

Since this was hard to read, I added the line to get the excel file, this looks like this;

x   
1   rs7079041
2   rs7244261
3   rs10448044
4   rs2610025
5   rs1472147
6   rs2648708
7   rs11891
8   rs1801725
9   rs6852678
10  rs3135758
11  rs6838240
12  rs6838240
13  rs603894
14  rs3764796
15  rs3764796
16  rs2073214
17  rs4971100
18  rs4971100
19  rs11718502
20  rs10888073
21  rs7032317

I also found another possible solution on here, but I again only got a list of my SNPs.

rows.diff <- function(catalog, proxy)
{
  catalogsnps.vec <- apply(catalogsnps, 1, paste, collapse="")
  proxy.vec <- apply(proxy, 1, paste, collaspse= "")
  rows.diff <- catalogsnps[!catalogsnps.vec %in% proxy.vec,]
  return(rows.diff)
}
write.table(rows.diff(catalogsnps, proxy), file="rowdiff.csv", sep=";", dec=",")

For the second part of my question, I am completely lost about where to start

Many thanks for any help

Claire

Answer 1

Why not just:

new.data <- merge(snap, catalog, by.x='proxy', by.y='catalogsnps')

This should give you a new dataframe whose rows are only the rows where proxy and catalogsnps match, and whose columns include all of the columns from the original dataframes.