The mapping rate is low in version 0.14.1, but other versions are normal

[gsx-ucas]

Hi there, I'am using a old version (0.14.1) of salmon and often get a low mapping rate of my samples. This time I got the lowest ratio ever, 0.0417748%.

[2021-04-16 16:40:11.804] [jointLog] [info] Fragment incompatibility prior below threshold.  Incompatible fragments will be ig
nored.
[2021-04-16 16:40:11.804] [jointLog] [info] Usage of --validateMappings implies use of minScoreFraction. Since not explicitly
specified, it is being set to 0.65
[2021-04-16 16:40:11.804] [jointLog] [info] Usage of --validateMappings, without --hardFilter implies use of range factorizati
on. rangeFactorizationBins is being set to 4
[2021-04-16 16:40:11.804] [jointLog] [info] Usage of --validateMappings implies a default consensus slack of 0.2. Setting cons
ensusSlack to 0.2.
[2021-04-16 16:40:11.804] [jointLog] [info] parsing read library format
[2021-04-16 16:40:11.804] [jointLog] [info] There is 1 library.
[2021-04-16 16:40:11.898] [jointLog] [info] Loading Quasi index
[2021-04-16 16:40:11.898] [jointLog] [info] Loading 32-bit quasi index
[2021-04-16 16:40:17.838] [jointLog] [info] done
[2021-04-16 16:40:17.838] [jointLog] [info] Index contained 77,654 targets
[2021-04-16 16:57:15.485] [fileLog] [info]
At end of round 0

Observed 25613538 total fragments (25613538 in most recent round)

[2021-04-16 16:57:15.485] [jointLog] [info] Computed 1,663 rich equivalence classes for further processing
[2021-04-16 16:57:15.485] [jointLog] [info] Counted 10,700 total reads in the equivalence classes
[2021-04-16 16:57:15.488] [jointLog] [warning] 0.00162414% of fragments were shorter than the k used to build the index (21).
If this fraction is too large, consider re-building the index with a smaller k.
The minimum read size found was 15.


[2021-04-16 16:57:15.488] [jointLog] [info] Number of mappings discarded because of alignment score : 86,602
[2021-04-16 16:57:15.488] [jointLog] [info] Number of fragments entirely discarded because of alignment score : 124,571
[2021-04-16 16:57:15.488] [jointLog] [info] Number of fragments discarded because they are best-mapped to decoys : 0
[2021-04-16 16:57:15.488] [jointLog] [info] Number of fragments discarded because they have only dovetail (discordant) mappings : 32,784
[2021-04-16 16:57:15.493] [jointLog] [warning] Only 10700 fragments were mapped, but the number of burn-in fragments was set to 5000000.
The effective lengths have been computed using the observed mappings.

[2021-04-16 16:57:15.493] [jointLog] [info] Mapping rate = 0.0417748%

[2021-04-16 16:57:15.493] [jointLog] [info] finished quantifyLibrary()
[2021-04-16 16:57:15.535] [jointLog] [info] Starting optimizer
[2021-04-16 16:57:15.564] [jointLog] [info] Marked 0 weighted equivalence classes as degenerate
[2021-04-16 16:57:15.568] [jointLog] [info] iteration = 0 | max rel diff. = 6.66952
[2021-04-16 16:57:16.080] [jointLog] [info] iteration = 100 | max rel diff. = 0.0161125
[2021-04-16 16:57:16.555] [jointLog] [info] iteration = 194 | max rel diff. = 0.000157223
[2021-04-16 16:57:16.558] [jointLog] [info] Finished optimizer
[2021-04-16 16:57:16.558] [jointLog] [info] writing output

[2021-04-16 16:57:16.681] [jointLog] [info] Computing gene-level abundance estimates
[2021-04-16 16:57:16.775] [jointLog] [info] There were 77689 transcripts mapping to 26673 genes
[2021-04-16 16:57:16.775] [jointLog] [info] NOTE: We recommend using tximport (https://bioconductor.org/packages/release/bioc/html/tximport.html) for aggregating transcript-level salmon abundance estimates to the gene level.  It is more versatile, exposes more features, and allows considering multi-sample information during aggregation.
[2021-04-16 16:57:16.911] [jointLog] [info] Aggregating expressions to gene level
[2021-04-16 16:57:17.077] [jointLog] [info] done
[2021-04-16 16:57:17.088] [jointLog] [warning] NOTE: Read Lib [[ Trim/1-2-intestines-LDC4673.filtered.R1.fq.gz, Trim/1-2-intestines-LDC4673.filtered.R2.fq.gz]] :

Detected a *potential* strand bias > 1% in an unstranded protocol check the file: Salmon_out/1-2-intestines-LDC4673/lib_format_counts.json for details

However，when I try old version (0.9.1) or latest version (1.4.0), I got a normal mapping rate (77.1074%):

[2021-04-20 10:08:58.047] [jointLog] [info] setting maxHashResizeThreads to 10
[2021-04-20 10:08:58.048] [jointLog] [info] Fragment incompatibility prior below threshold.  Incompatible fragments will be ignored.
[2021-04-20 10:08:58.048] [jointLog] [info] Usage of --validateMappings implies use of minScoreFraction. Since not explicitly specified, it is being set to 0.65
[2021-04-20 10:08:58.048] [jointLog] [info] Setting consensusSlack to selective-alignment default of 0.35.
[2021-04-20 10:08:58.048] [jointLog] [info] parsing read library format
[2021-04-20 10:08:58.048] [jointLog] [info] There is 1 library.
[2021-04-20 10:08:58.145] [jointLog] [info] Loading pufferfish index
[2021-04-20 10:08:58.146] [jointLog] [info] Loading dense pufferfish index.
[2021-04-20 10:09:05.198] [jointLog] [info] done
[2021-04-20 10:09:05.198] [jointLog] [info] Index contained 77,653 targets
[2021-04-20 10:09:05.266] [jointLog] [info] Number of decoys : 0
[2021-04-20 10:09:05.902] [jointLog] [info] Automatically detected most likely library type as IU

[2021-04-20 10:11:39.181] [fileLog] [info]
At end of round 0

Observed 25525591 total fragments (25525591 in most recent round)

[2021-04-20 10:11:39.181] [jointLog] [info] Computed 111,080 rich equivalence classes for further processing
[2021-04-20 10:11:39.181] [jointLog] [info] Counted 19,682,132 total reads in the equivalence classes
[2021-04-20 10:11:39.190] [jointLog] [warning] 0.00192356% of fragments were shorter than the k used to build the index.
If this fraction is too large, consider re-building the index with a smaller k.
The minimum read size found was 15.


[2021-04-20 10:11:39.190] [jointLog] [info] Number of mappings discarded because of alignment score : 17,018,803
[2021-04-20 10:11:39.190] [jointLog] [info] Number of fragments entirely discarded because of alignment score : 1,750,387
[2021-04-20 10:11:39.190] [jointLog] [info] Number of fragments discarded because they are best-mapped to decoys : 0
[2021-04-20 10:11:39.190] [jointLog] [info] Number of fragments discarded because they have only dovetail (discordant) mappings to valid targets : 234,041
[2021-04-20 10:11:39.190] [jointLog] [info] Mapping rate = 77.1074%

[2021-04-20 10:11:39.190] [jointLog] [info] finished quantifyLibrary()
[2021-04-20 10:11:39.193] [jointLog] [info] Starting optimizer
[2021-04-20 10:11:39.263] [jointLog] [info] Marked 0 weighted equivalence classes as degenerate
[2021-04-20 10:11:39.273] [jointLog] [info] iteration = 0 | max rel diff. = 14720
[2021-04-20 10:11:40.059] [jointLog] [info] iteration = 100 | max rel diff. = 11.9398
[2021-04-20 10:11:40.838] [jointLog] [info] iteration = 200 | max rel diff. = 8.95043
[2021-04-20 10:11:41.616] [jointLog] [info] iteration = 300 | max rel diff. = 8.965
[2021-04-20 10:11:42.390] [jointLog] [info] iteration = 400 | max rel diff. = 2.77483
[2021-04-20 10:11:43.170] [jointLog] [info] iteration = 500 | max rel diff. = 18.7693
[2021-04-20 10:11:43.921] [jointLog] [info] iteration = 600 | max rel diff. = 2.07666
[2021-04-20 10:11:44.677] [jointLog] [info] iteration = 700 | max rel diff. = 0.395157
[2021-04-20 10:11:45.418] [jointLog] [info] iteration = 800 | max rel diff. = 0.905927
[2021-04-20 10:11:46.136] [jointLog] [info] iteration = 900 | max rel diff. = 1.75975
[2021-04-20 10:11:46.842] [jointLog] [info] iteration = 1,000 | max rel diff. = 2.40606
[2021-04-20 10:11:47.548] [jointLog] [info] iteration = 1,100 | max rel diff. = 0.715333
[2021-04-20 10:11:48.250] [jointLog] [info] iteration = 1,200 | max rel diff. = 0.181266
[2021-04-20 10:11:48.938] [jointLog] [info] iteration = 1,300 | max rel diff. = 0.152511
[2021-04-20 10:11:49.636] [jointLog] [info] iteration = 1,400 | max rel diff. = 0.11611
[2021-04-20 10:11:50.315] [jointLog] [info] iteration = 1,500 | max rel diff. = 0.482892
[2021-04-20 10:11:51.013] [jointLog] [info] iteration = 1,600 | max rel diff. = 0.0388567
[2021-04-20 10:11:51.673] [jointLog] [info] iteration = 1,700 | max rel diff. = 0.518366
[2021-04-20 10:11:52.333] [jointLog] [info] iteration = 1,800 | max rel diff. = 0.117083
[2021-04-20 10:11:53.007] [jointLog] [info] iteration = 1,900 | max rel diff. = 0.306961
[2021-04-20 10:11:53.694] [jointLog] [info] iteration = 2,000 | max rel diff. = 0.078216
[2021-04-20 10:11:54.354] [jointLog] [info] iteration = 2,100 | max rel diff. = 1.17575
[2021-04-20 10:11:54.521] [jointLog] [info] iteration = 2,125 | max rel diff. = 0.00514978
[2021-04-20 10:11:54.541] [jointLog] [info] Finished optimizer
[2021-04-20 10:11:54.541] [jointLog] [info] writing output

[2021-04-20 10:11:54.750] [jointLog] [info] Computing gene-level abundance estimates
[2021-04-20 10:11:54.845] [jointLog] [info] There were 77689 transcripts mapping to 26673 genes
[2021-04-20 10:11:54.846] [jointLog] [info] NOTE: We recommend using tximport (https://bioconductor.org/packages/release/bioc/html/tximport.html) for aggregating transcript-level salmon abundance estimates to the gene level.  It is more versatile, exposes more features, and allows considering multi-sample information during aggregation.
[2021-04-20 10:11:55.040] [jointLog] [info] Aggregating expressions to gene level
[2021-04-20 10:11:55.126] [jointLog] [info] done

So, are there some bugs in version 0.14.1 ?

[rob-p]

There is not a known bug here, but this is certainly strange. Given that version 0.14.1 is very old now, I think the best thing to do is just put a caveat on the release page.