search

zhaoyanswill / RAPSearch2

Reduced Alphabet based Protein similarity Search
40 stars 15 forks source link

weird e-value #37

Open lmdu opened 7 years ago

lmdu commented 7 years ago

Hi, I used rapsearch2 to align DNA sequence to swissprot database, but I got weird evalues that do not like blast evalue.

TCONS_00000002 F4J9M5.1 41.2 97 56 1 3140 3427 497 593 0x1.08cb95295ce89p-303 73.6 TCONS_00000002 F4J9M5.1 51.0 157 77 0 3782 4252 708 864 0x1.08cb95295ce89p-303 168.7 TCONS_00000002 Q7ZU90.1 54.5 156 69 1 3782 4249 331 484 0x1.0e2cdd35f73f5p-300 166.8 TCONS_00000002 Q7ZU90.1 83.3 6 1 0 12592 12609 837 842 0x1.0e2cdd35f73f5p-300 16.5 TCONS_00000002 Q7ZU90.1 43.0 93 51 2 2831 3103 35 127 0x1.0e2cdd35f73f5p-300 75.9 TCONS_00000002 Q7ZU90.1 39.5 190 110 3 3155 3715 134 321 0x1.0e2cdd35f73f5p-300 144.4 TCONS_00000002 P41877.2 46.6 88 46 1 3452 3715 333 419 0x1.aa8e497595dc8p-298 87.4 TCONS_00000002 P41877.2 49.4 178 85 2 3752 4279 423 597 0x1.aa8e497595dc8p-298 171.4 TCONS_00000002 P41877.2 51.0 51 25 0 2798 2950 121 171 0x1.aa8e497595dc8p-298 57.0 TCONS_00000002 P41877.2 45.2 93 51 0 3155 3433 231 323 0x1.aa8e497595dc8p-298 85.1 TCONS_00000002 Q9CXF7.1 55.8 156 67 1 3782 4249 336 489 0x1.976a00d8b11fep-297 170.2 TCONS_00000002 Q9CXF7.1 43.8 96 52 2 2822 3103 37 132 0x1.976a00d8b11fep-297 80.9 TCONS_00000002 Q9CXF7.1 37.4 190 112 4 3158 3715 140 326 0x1.976a00d8b11fep-297 127.9

zhaoyanswill commented 6 years ago

Hi,

RAPSearch2 outputs log(E-value) by default. Thanks!

On Sat, Oct 14, 2017 at 2:40 AM, Lianming Du notifications@github.com wrote:

Hi, I used rapsearch2 to align DNA sequence to swissprot database, but I got weird evalues that do not like blast evalue.

TCONS_00000002 F4J9M5.1 41.2 97 56 1 3140 3427 497 593 0x1.08cb95295ce89p-303 73.6 TCONS_00000002 F4J9M5.1 51.0 157 77 0 3782 4252 708 864 0x1.08cb95295ce89p-303 168.7 TCONS_00000002 Q7ZU90.1 54.5 156 69 1 3782 4249 331 484 0x1.0e2cdd35f73f5p-300 166.8 TCONS_00000002 Q7ZU90.1 83.3 6 1 0 12592 12609 837 842 0x1.0e2cdd35f73f5p-300 16.5 TCONS_00000002 Q7ZU90.1 43.0 93 51 2 2831 3103 35 127 0x1.0e2cdd35f73f5p-300 75.9 TCONS_00000002 Q7ZU90.1 39.5 190 110 3 3155 3715 134 321 0x1.0e2cdd35f73f5p-300 144.4 TCONS_00000002 P41877.2 46.6 88 46 1 3452 3715 333 419 0x1.aa8e497595dc8p-298 87.4 TCONS_00000002 P41877.2 49.4 178 85 2 3752 4279 423 597 0x1.aa8e497595dc8p-298 171.4 TCONS_00000002 P41877.2 51.0 51 25 0 2798 2950 121 171 0x1.aa8e497595dc8p-298 57.0 TCONS_00000002 P41877.2 45.2 93 51 0 3155 3433 231 323 0x1.aa8e497595dc8p-298 85.1 TCONS_00000002 Q9CXF7.1 55.8 156 67 1 3782 4249 336 489 0x1.976a00d8b11fep-297 170.2 TCONS_00000002 Q9CXF7.1 43.8 96 52 2 2822 3103 37 132 0x1.976a00d8b11fep-297 80.9 TCONS_00000002 Q9CXF7.1 37.4 190 112 4 3158 3715 140 326 0x1.976a00d8b11fep-297 127.9

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/zhaoyanswill/RAPSearch2/issues/37, or mute the thread https://github.com/notifications/unsubscribe-auth/AETkp_lG_12DKX9yD1rKofapB5vwkY0mks5ssFdggaJpZM4P5R-v .

-- Sincerely, Y. Z.

pcantalupo commented 6 years ago

Hi I have been using Rapsearch2 v2.24 on a supercomputer (Linux) for several years. I use the '-s f' option to get regular evalues.

Now, I tried to set up Rapsearch2 v2.24 on a different machine (Ubuntu 16.04 VM). I compiled Rapsearch2 successfully and tried running it with '-s f' option. I see the same weird evalues as @lmdu.

Why is this happening? Thank you

edit: here is the output

zhaoyanswill commented 6 years ago

Hi,

This is strange. Does it happen on other Ubuntu machine? Thanks!

On Tue, Nov 7, 2017 at 1:55 PM, Paul Cantalupo notifications@github.com wrote:

Hi I have been using Rapsearch2 v2.24 on a supercomputer (Linux) for several years. I use the '-s f' option to get regular evalues.

Now, I tried to set up Rapsearch2 v2.24 on a different machine (Ubuntu 16.04 VM). I compiled Rapsearch2 successfully and tried running it with '-s f' option. I see the same weird evalues as @lmdu https://github.com/lmdu .

Why is this happening? Thank you

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/zhaoyanswill/RAPSearch2/issues/37#issuecomment-342584855, or mute the thread https://github.com/notifications/unsubscribe-auth/AETkpyW9Y82OYjTnBW2Hw4LvlPj8OcTbks5s0KeygaJpZM4P5R-v .

-- Sincerely, Y. Z.

pcantalupo commented 6 years ago

A colleague of mine fixed the problem with the following changes

He wrote to me:

OK, after some debugging looks like interpretation of formatting flags in C++ has changed. See here in C++11 section and compare with C++98 http://www.cplusplus.com/reference/ios/fixed/ In this case the code sets fixed and scientific formatting. With newer C++ standard fixed + scientific => hexfloat format that I see for the first time J

RAPSearch2.24_64bits.formatting.patch.txt

zhaoyanswill commented 6 years ago

Hi,

Thanks for the patch! Good to know about this change!

On Tue, Nov 7, 2017 at 8:57 PM, Paul Cantalupo notifications@github.com wrote:

A colleague of mine fixed the problem with the following changes

He wrote to me:

OK, after some debugging looks like interpretation of formatting flags in C++ has changed. See here in C++11 section and compare with C++98 http://www.cplusplus.com/reference/ios/fixed/ In this case the code sets fixed and scientific formatting. With newer C++ standard fixed + scientific => hexfloat format that I see for the first time J

RAPSearch2.24_64bits.formatting.patch.txt https://github.com/zhaoyanswill/RAPSearch2/files/1452341/RAPSearch2.24_64bits.formatting.patch.txt

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/zhaoyanswill/RAPSearch2/issues/37#issuecomment-342684948, or mute the thread https://github.com/notifications/unsubscribe-auth/AETkp18pGzSC_l7JETHdtvYjn0PTOPU0ks5s0Qp4gaJpZM4P5R-v .

-- Sincerely, Y. Z.

metageni commented 6 years ago

@zhaoyanswill Any chance you can add this fix to rapsearch2?