Hai everybody,
I have been writing a small web based DAS client for some of our users. I have been building it on top of biojava. We are still on ensembl326 in house so everything here goes for the ensembl326 release. I have not run any tests against ensembl428 because the sanger servlet site on port 8080 is blocked by our firewall (any changes on getting it on port 80?).
Everything seemed to work fine, but testing my results against the inhouse ensembl326 website gives me right results in one area of the genome and wrong results in others.
I have tried to locate the problem, and used the TestDASG class in biojava to locate where the problem occurs. Attached is my modified version of TestDASG.java. I am running it against a standard ensembl326 DAS reference server. An example:
running TestDASG against 1 1 50000 gives the same result as ensembl: it maps in clone AL589746, and the transcripts show in the same place. However, running against 1 92950487 92970487 gives the wrong result. According to ensembl this maps to AL049597. TestDASG says it maps to AL353627. Ensembl gives the coordinates of AL353627 at arround 93.81 Mb, a completely different location...
I have been running against several areas, and the problem recurs. I use the standard biojava API's, I have looked at the raw XML result files, but I can not see where the problem occurs. I suspect there are problems with the orientation, but I am not really sure what is happening.
Can anybody give me some clues as to what I am doing wrong and how I can fix it? I'll probably spent some time this weekend to test against public servers to see if the problem occurs there too.
Frank
TestDASG.java
Description: Binary data
