Extracting values from IRanges objects in R/Bioconductor

Question

I've imported a UCSC alignability track into R using import.bw() (from the rtracklayer package) but am having trouble accessing the values I need.

For example: I want to provide a chromosome and a base and return the value at that position.

My object is called al100:

> al100
RangedData with 21591667 rows and 1 value column across 25 spaces
            space               ranges   |       score
         <factor>            <IRanges>   |   <numeric>
1            chr1       [10001, 10014]   | 0.002777778
2            chr1       [10015, 10015]   | 0.333333343
3            chr1       [10016, 10026]   | 0.500000000
4            chr1       [10027, 10031]   | 1.000000000

I want a function where I specify a chrosome and position and get back the score. This is trivial if I want one or two values, but a loop isn't going to work when I've got 7 million to look up; at 4/5 seconds per query, that's about 10 months, which is not an option.

For example, chr1, position 10011 would return the value 0.002777778 (where x is a separate object containing a list of chromosomes and positions)

The only method I've found so far is to ask if my position is equal or greater than the start and/or equal or equal to or less than the end of a range. Not very good.

score(al100["chr1"])[ which( start(al100["chr1"]<=x$POS[1])) & end(al100["chr1"]<=x$POS[1]))   ]

Answer 1

For a reproducible example

library(rtracklayer)
example(import.bw)
gffRD

gives

> head(gffRD, 3)
RangedData with 3 rows and 7 value columns across 1 space
                                  space       ranges |     type       source
                               <factor>    <IRanges> | <factor>     <factor>
1 Escherichia_coli_K-12_complete_genome [ 337, 2799] |      CDS glimmer/tico
2 Escherichia_coli_K-12_complete_genome [2801, 3733] |      CDS glimmer/tico
3 Escherichia_coli_K-12_complete_genome [3734, 5020] |      CDS glimmer/tico
     phase   strand        note     shift     score
  <factor> <factor> <character> <numeric> <numeric>
1       NA        +          NA        NA  5.347931
2       NA        +          NA        NA 11.448764
3       NA        +          NA        NA  6.230648

Define regions of interest

roi <- GRanges("Escherichia_coli_K-12_complete_genome",
               IRanges(c(337, 3734), width=1))

then use findOverlaps to map between gffRD and roi

olaps <- findOverlaps(gffRD,roi)
df <- DataFrame(seqnames=seqnames(roi)[subjectHits(olaps)],
                 start=start(roi)[subjectHits(olaps)],
                 Score=score(gffRD)[queryHits(olaps)])

olaps contains information about which queries match which subjects

> olaps
Hits of length 2
queryLength: 14
subjectLength: 2
  queryHits subjectHits 
   <integer>   <integer> 
 1         1           1 
 2         3           2 

The data frame is

> df
DataFrame with 2 rows and 3 columns
                               seqnames     start     Score
                                  <Rle> <integer> <numeric>
1 Escherichia_coli_K-12_complete_genome       337  5.347931
2 Escherichia_coli_K-12_complete_genome      3734  6.230648