Resume after skipQuant?

[xuesoso]

Hi, I have executed salmon with "--skipQuant". Is there a way to run just the quantification step without re-running the first phase of the algorithm?

Thanks!

[rob-p]

cc : @k3yavi

Hi @xuesoso, I know that Avi (tagged above) has implemented a flag for this, but I'm not certain if it is exposed in the current release. I'm tagging him here to chime in.

[k3yavi]

Hi @xuesoso ,

If you have dumped the salmon's rich equivalence classes in the first round then you can run salmon in alignment mode by specifying the path of the file with -e <File_PATH> instead of the BAM file for the requantification.

[xuesoso]

@k3yavi @rob-p Thank you for the prompt responses! I executed the following but seems to get no "quant.sf" as output (in fact no other output except for the log file). I don't find "AlignmentLibrary" object even though the log states that it did. Do you know what went wrong? Thank you!

Command:

salmon quant -e ./aux_info/eq_classes.txt.gz --libType IU -o ./

Output log: [2020-09-18 20:01:55.879] [jointLog] [info] setting maxHashResizeThreads to 8 [2020-09-18 20:01:55.879] [jointLog] [info] Fragment incompatibility prior below threshold. Incompatible fragments will be ignored. [2020-09-18 20:01:55.879] [jointLog] [info] numQuantThreads = 4 [2020-09-18 20:02:50.408] [jointLog] [warning] Missing effective lens for 47121 transcripts; setting to 100.0. [2020-09-18 20:02:50.408] [jointLog] [warning] NOTE: Since effective lengths are not provided, please do not rely on the TPM field in the ouput quantifications. Only the NumReads field will be reliable. [2020-09-18 20:02:50.410] [jointLog] [info] Found total 187671 eqclasses and 47121 transcripts [2020-09-18 20:02:50.682] [jointLog] [info] Created AlignmentLibrary object

My Salmon version is v1.3.0.