IDP - Isoform Detection and Prediction - new version 0.1.6

g3n3 · 08-04-2014, 01:45 PM

IDP is an gene Isoform Detection and Prediction tool from Second Generation Sequencing and PacBio sequencing.

It offers very reliable gene isoform identification with high sensitivity. This release includes both bug fixes and new features.

Now version 0.1.6 has been released. It offers some important additional features such as the option to use either GMAP or BLAT for long read alignment. It has also been released under the very open Apache 2.0 license.

You may find IDP here:
http://www.healthcare.uiowa.edu/labs/au/IDP/

If you have any questions please contact us.
http://www.healthcare.uiowa.edu/labs...DP_contact.asp

Release Notes:
1.) Bug fix: Now allowing computation of detected_exp_len file when not provided.
2.) estimator_choice being added to the config file and accepting
either MAP or MLE (MAP is default if the parameter is not set or
absent).
3.) Software is licensed under Apache 2.0
4.) GMAP can optionally be used as well as BLAT, however when using
GMAP, a path to the index must be specified in the config file
5.) BLAT and SeqMap are no longer bundled, paths to aligner
executables must be specified if they are not installed in the path
under their default names.

lingling huang · 06-02-2017, 12:22 AM

I used the software IDP to analyze iso-seq data. In the final results, I want to extract read sequences according to novel isoforms name from fasta file.
However, I can't find the corresponding transcript ID like‘ PB.1.1’ of the isoform name from IDP output results since the isoform name is the gene locus with an index (such as chr2:204193002-204296891.1).
Could you help me? I'd really appreciated that.

g3n3 · 06-02-2017, 08:09 AM

seqanswers lost my quick reply that had a long from answer.

basically I think you need to look in isoform.gpd output. I think you may be looking in a different file. There the known detected transcripts and genes have their appropriate names and annotations in gpd format

https://www.healthcare.uiowa.edu/lab...gpd_format.asp

novel transcripts of known genes get the "chr:..." ID attached to a known gene id.

novel genes get the "chr:..." ID attached to both transcript and gene.

This behavior is sufficient for tracking whats novel. However the major caveat to all this is that IDP's calls for what is a 'novel' are VERY sensitive, and sometimes transcripts called novel may be just barely (ie 1bp) different from a known transcript. The lion's share of things that are called novel, in my experience fall into this class of VERY similar transcripts, and for most practical use require a lot of scrutiny as to whether they are really novel or not, so I recommend loading your output in UCSC genome browser to see if your novel transcripts are what you are looking for.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to convert high version CEL file to low version?	lee_sh	Bioinformatics	1	05-23-2012 02:47 PM
cufflinks 1.2.0 version got me significantly different results than the old version	slowsmile	Bioinformatics	9	02-01-2012 02:26 AM
RNA-Seq: Isoform-level microRNA-155 target prediction using RNA-seq.	Newsbot!	Literature Watch	0	02-15-2011 03:00 AM
Newbler sensitivity tuning for isoform detection?	yvan.wenger	454 Pyrosequencing	2	12-07-2010 12:35 AM

08-04-2014, 01:45 PM	#1
g3n3 Junior Member Location: Hawaii Join Date: Jan 2013 Posts: 6	IDP - Isoform Detection and Prediction - new version 0.1.6 IDP is an gene Isoform Detection and Prediction tool from Second Generation Sequencing and PacBio sequencing. It offers very reliable gene isoform identification with high sensitivity. This release includes both bug fixes and new features. Now version 0.1.6 has been released. It offers some important additional features such as the option to use either GMAP or BLAT for long read alignment. It has also been released under the very open Apache 2.0 license. You may find IDP here: http://www.healthcare.uiowa.edu/labs/au/IDP/ If you have any questions please contact us. http://www.healthcare.uiowa.edu/labs...DP_contact.asp Release Notes: 1.) Bug fix: Now allowing computation of detected_exp_len file when not provided. 2.) estimator_choice being added to the config file and accepting either MAP or MLE (MAP is default if the parameter is not set or absent). 3.) Software is licensed under Apache 2.0 4.) GMAP can optionally be used as well as BLAT, however when using GMAP, a path to the index must be specified in the config file 5.) BLAT and SeqMap are no longer bundled, paths to aligner executables must be specified if they are not installed in the path under their default names.

06-02-2017, 12:22 AM	#2
lingling huang Member Location: changsha Join Date: Mar 2016 Posts: 46	How to find the corresponding name of isoform name is the gene locus with an index (s I used the software IDP to analyze iso-seq data. In the final results, I want to extract read sequences according to novel isoforms name from fasta file. However, I can't find the corresponding transcript ID like‘ PB.1.1’ of the isoform name from IDP output results since the isoform name is the gene locus with an index (such as chr2:204193002-204296891.1). Could you help me? I'd really appreciated that.

06-02-2017, 08:09 AM	#3
g3n3 Junior Member Location: Hawaii Join Date: Jan 2013 Posts: 6	seqanswers lost my quick reply that had a long from answer. basically I think you need to look in isoform.gpd output. I think you may be looking in a different file. There the known detected transcripts and genes have their appropriate names and annotations in gpd format https://www.healthcare.uiowa.edu/lab...gpd_format.asp novel transcripts of known genes get the "chr:..." ID attached to a known gene id. novel genes get the "chr:..." ID attached to both transcript and gene. This behavior is sufficient for tracking whats novel. However the major caveat to all this is that IDP's calls for what is a 'novel' are VERY sensitive, and sometimes transcripts called novel may be just barely (ie 1bp) different from a known transcript. The lion's share of things that are called novel, in my experience fall into this class of VERY similar transcripts, and for most practical use require a lot of scrutiny as to whether they are really novel or not, so I recommend loading your output in UCSC genome browser to see if your novel transcripts are what you are looking for.