0001555: Problem with the program using MacBook Air

(0001350)
user19
1969-12-31 17:33

Data type = Nucleotide
Protein Coding = NO
File = Attached

Error:
Type = #/0
File = MLModels.pas
Object = TGTRModel
Function = SetParamsFromSeqs
Line = Fi[i] := 1/sqrt(FFreq[i])

(0001355)
agasky (reporter)
1969-12-31 17:33

Dan, since you commented on this an apparently investigated it, please assign the bug to yourself.

(0001365)
Nikita Vikhrev (reporter)
1969-12-31 17:33

This is not easy to be fixed. I am not sure whether GTR model is applicable to this data. I will think about it.

(0001368)
user19
1969-12-31 17:33

The problem turned out to be that MEGA wasn\'t supporting RNA. In the attached file there were no \'T\' characters, only \'U\'. Since it counted 0 \'T\' characters and we eventually got to a line where we divided by the frequency of \'T\' (which is 0) we got a div by 0 error.

I added support so that \'U\' is threated the same as \'T\'. We still need to fix the function SetParamsFromSeqs so that if we have 0 \'A\', \'T or U\', \'C\', or \'G\' then it doesn\'t continue and cause a div by 0 error.

(0001375)
Nikita Vikhrev (reporter)
1969-12-31 17:33

Yet, we have to support the cases where frequencies of some nucleotide is really zero.

(0001376)
Nikita Vikhrev (reporter)
1969-12-31 17:33

I have solve the problem all. I will send you the code after I return (I am out of town now).

(0004331)
gstecher (administrator)
2019-11-11 11:21

Hi Matheus,

I am writing in response to your question regarding the MEGA software. If you are having issues running MEGA X on your mac system, you could try downloading the MEGA7 version for Windows and running it in a virtual machine (you could use something like VMWare Fusion or Parallels).

--
Best regards,
Glen Stecher
Institute for Genomics and Evolutionary Medicine
igem.temple.edu

Issue History
Date Modified	Username	Field	Change
1969-12-31 17:33	Nikita Vikhrev	Status	assigned => resolved
1969-12-31 17:33	Nikita Vikhrev	Resolution	open => fixed
2019-11-10 22:43	guest	New Issue
2019-11-11 10:04	gstecher	File Deleted: mle.zip
2019-11-11 11:21	gstecher	Note Added: 0004331
2019-11-11 11:21	gstecher	Status	new => resolved
2019-11-11 11:21	gstecher	Resolution	open => unable to reproduce
2019-11-11 11:21	gstecher	Assigned To	=> gstecher

Notes
(0001350) user19 1969-12-31 17:33	Data type = Nucleotide Protein Coding = NO File = Attached Error: Type = #/0 File = MLModels.pas Object = TGTRModel Function = SetParamsFromSeqs Line = Fi[i] := 1/sqrt(FFreq[i])

(0001355) agasky (reporter) 1969-12-31 17:33	Dan, since you commented on this an apparently investigated it, please assign the bug to yourself.

(0001365) Nikita Vikhrev (reporter) 1969-12-31 17:33	This is not easy to be fixed. I am not sure whether GTR model is applicable to this data. I will think about it.

(0001368) user19 1969-12-31 17:33	The problem turned out to be that MEGA wasn\'t supporting RNA. In the attached file there were no \'T\' characters, only \'U\'. Since it counted 0 \'T\' characters and we eventually got to a line where we divided by the frequency of \'T\' (which is 0) we got a div by 0 error. I added support so that \'U\' is threated the same as \'T\'. We still need to fix the function SetParamsFromSeqs so that if we have 0 \'A\', \'T or U\', \'C\', or \'G\' then it doesn\'t continue and cause a div by 0 error.

(0001375) Nikita Vikhrev (reporter) 1969-12-31 17:33	Yet, we have to support the cases where frequencies of some nucleotide is really zero.

(0001376) Nikita Vikhrev (reporter) 1969-12-31 17:33	I have solve the problem all. I will send you the code after I return (I am out of town now).

(0004331) gstecher (administrator) 2019-11-11 11:21	Hi Matheus, I am writing in response to your question regarding the MEGA software. If you are having issues running MEGA X on your mac system, you could try downloading the MEGA7 version for Windows and running it in a virtual machine (you could use something like VMWare Fusion or Parallels). -- Best regards, Glen Stecher Institute for Genomics and Evolutionary Medicine igem.temple.edu

Relationships