search

dstreett / FLASH2

Flash2 has some improvements from flash_1 including new logic from innie and outie overlaps as well as some initial steps for flash for amplicons
GNU General Public License v3.0
47 stars 12 forks source link

Master branch is broken #3

Open msettles opened 8 years ago

msettles commented 8 years ago

2 issues at least, first is a ton of what look like debugging print statement are output to screen, second produces a non-valid read that breaks everything.

R just dies, says it expects a blank line my python app says a read has more quality values than read characters.

msettles commented 8 years ago

Had to go back to the previous commit March 1, 2015 for it to work, this version contains the bug with respect to complete overlaps that are completely within the read, I remember something on this?

dstreett commented 8 years ago

What kind of non-valid read are you seeing? Is this preprocess? I ran a combination of both SE and PE reads through the pipeline and I'm not seeing any invalid reads. Is this under Brice's directory?

On Sat, Feb 6, 2016 at 10:58 AM, Matt Settles notifications@github.com wrote:

Had to go back to the previous commit March 1, 2015 for it to work, this version contains the bug with respect to complete overlaps that are completely within the read, I remember something on this?

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180838129.

msettles commented 8 years ago

I don't run anything on the Idaho machines anymore, this is from pulling down the head on github. Is github the most up to date? On Feb 6, 2016 11:30 AM, "David Streett" notifications@github.com wrote:

What kind of non-valid read are you seeing? Is this preprocess? I ran a combination of both SE and PE reads through the pipeline and I'm not seeing any invalid reads. Is this under Brice's directory?

On Sat, Feb 6, 2016 at 10:58 AM, Matt Settles notifications@github.com wrote:

Had to go back to the previous commit March 1, 2015 for it to work, this version contains the bug with respect to complete overlaps that are completely within the read, I remember something on this?

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180838129.

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180843626.

dstreett commented 8 years ago

Yes, I pushed a copy and paste bug. Testing now- making sure the change doesn't break anything else.

On Sat, Feb 6, 2016 at 11:50 AM, Matt Settles notifications@github.com wrote:

I don't run anything on the Idaho machines anymore, this is from pulling down the head on github. Is github the most up to date? On Feb 6, 2016 11:30 AM, "David Streett" notifications@github.com wrote:

What kind of non-valid read are you seeing? Is this preprocess? I ran a combination of both SE and PE reads through the pipeline and I'm not seeing any invalid reads. Is this under Brice's directory?

On Sat, Feb 6, 2016 at 10:58 AM, Matt Settles notifications@github.com wrote:

Had to go back to the previous commit March 1, 2015 for it to work, this version contains the bug with respect to complete overlaps that are completely within the read, I remember something on this?

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180838129.

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180843626.

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180846173.

dstreett commented 8 years ago

Would you mind checking out the development branch? I fixed the biggest issue and removed the print statements.

On Sat, Feb 6, 2016 at 11:50 AM, Matt Settles notifications@github.com wrote:

I don't run anything on the Idaho machines anymore, this is from pulling down the head on github. Is github the most up to date? On Feb 6, 2016 11:30 AM, "David Streett" notifications@github.com wrote:

What kind of non-valid read are you seeing? Is this preprocess? I ran a combination of both SE and PE reads through the pipeline and I'm not seeing any invalid reads. Is this under Brice's directory?

On Sat, Feb 6, 2016 at 10:58 AM, Matt Settles notifications@github.com wrote:

Had to go back to the previous commit March 1, 2015 for it to work, this version contains the bug with respect to complete overlaps that are completely within the read, I remember something on this?

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180838129.

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180843626.

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180846173.

msettles commented 8 years ago

Out on the town, but as soon as soon as I get back home On Feb 6, 2016 1:08 PM, "David Streett" notifications@github.com wrote:

Would you mind checking out the development branch? I fixed the biggest issue and removed the print statements.

On Sat, Feb 6, 2016 at 11:50 AM, Matt Settles notifications@github.com wrote:

I don't run anything on the Idaho machines anymore, this is from pulling down the head on github. Is github the most up to date? On Feb 6, 2016 11:30 AM, "David Streett" notifications@github.com wrote:

What kind of non-valid read are you seeing? Is this preprocess? I ran a combination of both SE and PE reads through the pipeline and I'm not seeing any invalid reads. Is this under Brice's directory?

On Sat, Feb 6, 2016 at 10:58 AM, Matt Settles < notifications@github.com> wrote:

Had to go back to the previous commit March 1, 2015 for it to work, this version contains the bug with respect to complete overlaps that are completely within the read, I remember something on this?

— Reply to this email directly or view it on GitHub <https://github.com/dstreett/FLASH2/issues/3#issuecomment-180838129 .

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180843626.

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180846173.

— Reply to this email directly or view it on GitHub https://github.com/dstreett/FLASH2/issues/3#issuecomment-180863922.

dstreett commented 8 years ago

Cool, thanks, Matt.

msettles commented 8 years ago

FLASH2 develop compiles are run properly, reads are good.

Statistics were the same as the March version (which they should be correct?). However the extended reads look more accurate to me than before, though there are 2 reads out of 19,298 which aren’t merged by my expectations. Specifically in this study every Read 1 begins with a TA (after adapter trimming), however since we define anything prior to the start of read 1 as adapter (correct?), then every reads post FLASH merged or not should still begin with a TA (these 2 do not). Turns out to be completely legitimate overlap and in this case the second base is in disagreement and read2 base is chosen, though I think its the lower quality base (See attachment). Need a second (or third) set of eyes, but here are two reads

@J00113:93:H5GNNBBXX:6:1101:26758:1894_TCGGAACA|Linker_Sol_AP1-3_0|GAAGCAAAT|Primer_Sol_Mar_1b_0_Adapter_2.2_Bar_B*1 1:N:0:TCGGAACA TACACCGGCGCCGCCGGCACCACCCGGACCGCCGGCAAAACCCTACCCGCCTTGCTTCCACGAACCACCGCGGCCA + AFFJJJJJJFJJJJJJ-7AJJ-FJ<-7-AA-AA<A-<---7---<F<-7--------------<AF--A--7-<--

@J00113:93:H5GNNBBXX:6:1101:26758:1894_TCGGAACA|Linker_Sol_AP1-3_0|GAAGCAAAT|Primer_Sol_Mar_1b_0_Adapter_2.2_Bar_B*1 2:N:0:TCGGAACA GCCGGGGGCGCCGGTGGTGTCGGCGGCGCCGGGGGAACAGGTGGGCTGA + JJJJJJJJJJJJJJJJJJJAFJJFJ7AJJJJJ-A-7-7-7<A-7AFFJA

@J00113:93:H5GNNBBXX:6:1101:22079:20304_TCGGAACA|Linker_Sol_AP1-1_0|CCATAACAC|Primer_Sol_Mar_5b_0_Adapter_2.2_Bar_B*0 1:N:0:TCGGAACA TAAGTGGTGGTCGCGGCGGGGCCGGCGGTAACGCCGGCACGTTCTACGGTTCCGGCGGCGCCGGCGGCGCCG + AJJJJJFJJFJJJJFJJJJJFFFFJJJJFJJJJFJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJFJJJJJJ

@J00113:93:H5GNNBBXX:6:1101:22079:20304_TCGGAACA|Linker_Sol_AP1-1_0|CCATAACAC|Primer_Sol_Mar_5b_0_Adapter_2.2_Bar_B*0 2:N:0:TCGGAACA AGGCGTTGCCGCCGTTACCGCCGCTACCACCCTTTCCTGCTATCGTGGCACCGC + JFJJJJJJJJJJJJJJJJJJJJJJJJJJJJJFJJJJJJFJFFFJJJ<JAJJJFJ

The flashed result GAAGTGGTGGTCGCGGCGGGAACGGCGGCAACGCCG

I can live with the 2 for sure and add another check for that TA downstream (or I for this work I may increase the min outie overlap). Matt

From: David Streett notifications@github.com Reply-To: dstreett/FLASH2 reply@reply.github.com Date: Saturday, February 6, 2016 at 7:32 PM To: dstreett/FLASH2 FLASH2@noreply.github.com Cc: Matt Settles mattsettles@gmail.com Subject: Re: [FLASH2] Master branch is broken (#3)

Cool, thanks, Matt.

— Reply to this email directly or view it on GitHub.

msettles commented 8 years ago

don’t forget to update the version number

From: David Streett notifications@github.com Reply-To: dstreett/FLASH2 reply@reply.github.com Date: Saturday, February 6, 2016 at 7:32 PM To: dstreett/FLASH2 FLASH2@noreply.github.com Cc: Matt Settles mattsettles@gmail.com Subject: Re: [FLASH2] Master branch is broken (#3)

Cool, thanks, Matt.

— Reply to this email directly or view it on GitHub.