Moderator: Michael Lawrence

gravatar for Michael Lawrence
Reputation:
8,830
Status:
Trusted
Location:
United States
Last seen:
4 days ago
Joined:
7 years, 1 month ago
Email:
l***************@gene.com

Posts by Michael Lawrence

<prev • 1,051 results • page 1 of 106 • next >
0
votes
3
answers
109
views
3
answers
Answer: A: Aggregate read counts by non-uniform overlapping sliding windows
... I guess the simple way would be to just generate all of the hits and aggregate the matrix by the subject hits: hits <- findOverlaps(queryRanges, subjectRanges, type="within") rowsum(countMatrix[queryHits(hits),], subjectHits(hits)) If that is too slow, we can work on it. If this was just a vec ...
written 4 days ago by Michael Lawrence8.8k
0
votes
2
answers
76
views
2
answers
Answer: A: Best way to create a VRanges object from a large VCF file?
... You really don't want to concatenate iteratively in this case, as there is a lot of overhead to each contatenation (and if you really want to iterate with side effects, use a for() loop for clarity). You should use lapply() when reading, not sapply(), since you really just want a list. And you can c ...
written 6 days ago by Michael Lawrence8.8k
0
votes
1
answers
61
views
1
answers
Comment: C: Why does a `first_time` object appear when showing GRangesList object?
... The comments do not indicate any rationale for this, as far as I can see. Looks like a regression due to 119047, which removed the protecting closure without changing <<- to <-.  I checked a fix for this into devel.   ...
written 6 days ago by Michael Lawrence8.8k
0
votes
1
answers
174
views
1
answers
Comment: C: `precede` does not perform as expected
... As I tried to say in the previous post (which I later edited), I think the contract of the function would be simpler if we just say that the orientation is determined by one of the arguments, default "subject". If the orientation argument has "*" strand, it means "none". Then, Janet's use case works ...
written 14 days ago by Michael Lawrence8.8k
0
votes
1
answers
174
views
1
answers
Comment: C: `precede` does not perform as expected
... I don't think a gene can follow/precede a peak, because a peak has no orientation. It's better to ask, which peaks precede (the transcription of) a gene, and then find the closest peak for each gene. The precede(), follow() interface is too low-level to make that simple. One option would be a high-l ...
written 14 days ago by Michael Lawrence8.8k
0
votes
1
answers
174
views
1
answers
Comment: C: `precede` does not perform as expected
... It would be nice to have a version of nearest() that does not allow overlaps. I discovered that when developing HelloRanges. Maybe a new verb, adjacent()? I kind of prefer that to a parameter like ignore.overlaps, just because it's simpler to list operations as functions, compared to function/parame ...
written 14 days ago by Michael Lawrence8.8k
0
votes
1
answers
174
views
1
answers
Comment: C: `precede` does not perform as expected
... I agree that a more explicit API can be a good idea. It makes it easier to document, and the code easier to read, up to a point. In this case though, I think your use and Janet's use case are both intuitively satisfied by ignoring strand when the subject has "*", not when x has "*". Adding options a ...
written 15 days ago by Michael Lawrence8.8k
0
votes
1
answers
76
views
1
answers
Answer: A: rtracklayer: how to view coordinates that failed liftover?
... liftOver() yields a GRangesList, with one element per input range. I'm not sure what you mean by "lines" but I guess you are unlisting the result and writing it to a file. Instead of unlisting, you can just look at the lengths() of the return value to determine the degree of mapping for each input r ...
written 15 days ago by Michael Lawrence8.8k
0
votes
1
answers
174
views
1
answers
Comment: C: `precede` does not perform as expected
... For the present use case, where the subject ranges have a strand, I agree that we should respect it.  It's not clear how the gene use case requires treating "*" in subject as "both". The strand has two functions in GenomicRanges: (1) specifying a subspace used in overlap detection (where ranges on o ...
written 15 days ago by Michael Lawrence8.8k
0
votes
1
answers
64
views
1
answers
Answer: A: count coverage for exact matches to sequence pattern
... For your findPattern() function, I suggest this implementation: hits <- GRanges(vmatchPattern(pat, mydnaseqset)) That assumes that gtfobject has names. You could provide some. Then, you can map the transcript-local ranges to global ranges like this: hits.global <- mapFromTranscripts(hits ...
written 15 days ago by Michael Lawrence8.8k

Latest awards to Michael Lawrence

Scholar 8 months ago, created an answer that has been accepted. For A: VariantAnnotation as.vranges with asterisk in ALT field
Scholar 9 months ago, created an answer that has been accepted. For A: VariantAnnotation as.vranges with asterisk in ALT field
Scholar 9 months ago, created an answer that has been accepted. For A: VariantAnnotation as.vranges with asterisk in ALT field
Scholar 9 months ago, created an answer that has been accepted. For A: VariantAnnotation as.vranges with asterisk in ALT field
Guru 9 months ago, received more than 100 upvotes.
Teacher 11 months ago, created an answer with at least 3 up-votes. For A: How to find the amino acid codon corresponding to a translated genomic position
Scholar 11 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 11 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 11 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 11 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 12 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 12 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 13 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Teacher 13 months ago, created an answer with at least 3 up-votes. For A: How to find the amino acid codon corresponding to a translated genomic position
Scholar 13 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 14 months ago, created an answer that has been accepted. For A: issue renaming seqnames in genomic ranges
Scholar 14 months ago, created an answer that has been accepted. For A: How to find particular types of overlap with GenomicRanges
Scholar 14 months ago, created an answer that has been accepted. For A: How to find particular types of overlap with GenomicRanges
Scholar 14 months ago, created an answer that has been accepted. For A: How to find particular types of overlap with GenomicRanges
Teacher 15 months ago, created an answer with at least 3 up-votes. For A: How to find particular types of overlap with GenomicRanges
Scholar 15 months ago, created an answer that has been accepted. For A: How to find particular types of overlap with GenomicRanges
Scholar 15 months ago, created an answer that has been accepted. For A: How to find particular types of overlap with GenomicRanges
Scholar 15 months ago, created an answer that has been accepted. For A: Can precede() and follow() in some cases return identical output?
Scholar 15 months ago, created an answer that has been accepted. For A: How to use GenomicRanges to find overlaps by percentage covered of interval
Scholar 15 months ago, created an answer that has been accepted. For A: importing bigwig file: "sequence names not known"

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.2.0
Traffic: 169 users visited in the last hour