Dear Alex,
Thanks for the workaround.
I tried a range of values (25,30,35,40,45) for --seedPerWindowNmax,
I ran on the same small data set, in rapid succession on the same computing resources.
Only 1-2 seconds for the mapping step using 25,30,35,40,45
Then the "blowup" using 50.
Now intellectually, I know:
"There is no "knee" in an exponential curve"
http://www.abarry.org/knee.htmBut it feels like a knee. Or maybe it's not precisely exponential growth.
No matter. Good workaround. Thanks again.
/Sol.
--------------------------------------------------------
====== seedPerWindow25 ====
Finished loading the genome: Tue Jul 18 10:19:07 2017
Jul 18 10:19:09 ..... started sorting BAM
====== seedPerWindow30 ====
Finished loading the genome: Tue Jul 18 10:32:34 2017
Jul 18 10:32:35 ..... started sorting BAM
====== seedPerWindow35 ====
Finished loading the genome: Tue Jul 18 10:33:02 2017
Jul 18 10:33:03 ..... started sorting BAM
====== seedPerWindow40 ====
Finished loading the genome: Tue Jul 18 10:33:30 2017
Jul 18 10:33:31 ..... started sorting BAM
====== seedPerWindow45 ====
Finished loading the genome: Tue Jul 18 10:33:58 2017
Jul 18 10:34:00 ..... started sorting BAM