Hi everyone,
I'm trying to install MACS3 via anaconda in order to do single-cell ATAC sequencing analysis via Signac in R.
I'm trying to install in Windows 10 which, it seems, is my first mistake, since I've heard from several people that it works very well in Linux but not with Windows. However, I don't have that option right now so I'm stuck with this.
I've already debugged the installation to a certain extent - I've installed gcc, then:
Had to catch a KeyError in the setup script after it looked for sysconfig.get_system_vars()['CC"], which with my system doesn't exist - I don't think this was super important as there was an elif statement below which my system could satisfy.
Got an error trying to attach the "m" library to extensions, which wasn't found - I just deleted these arguments
The setup script then ran, but threw errors when trying to run swalign.c:
MACS3/Signal/swalign.c(97): error C2057: Konstanter Ausdruck erwartet
MACS3/Signal/swalign.c(97): error C2466: Zuordnung eines Arrays der konstanten Größe 0 nicht möglich.
MACS3/Signal/swalign.c(97): error C2133: "reversed1": Unbekannte Größe
MACS3/Signal/swalign.c(98): error C2057: Konstanter Ausdruck erwartet
MACS3/Signal/swalign.c(98): error C2466: Zuordnung eines Arrays der konstanten Größe 0 nicht möglich.
MACS3/Signal/swalign.c(98): error C2133: "reversed2": Unbekannte Größe
MACS3/Signal/swalign.c(99): error C2057: Konstanter Ausdruck erwartet
MACS3/Signal/swalign.c(99): error C2466: Zuordnung eines Arrays der konstanten Größe 0 nicht möglich.
MACS3/Signal/swalign.c(99): error C2133: "reversed3": Unbekannte Größe
MACS3/Signal/swalign.c(339): error C2057: Konstanter Ausdruck erwartet
MACS3/Signal/swalign.c(339): error C2466: Zuordnung eines Arrays der konstanten Größe 0 nicht möglich.
MACS3/Signal/swalign.c(339): error C2133: "c": Unbekannte Größe
MACS3/Signal/swalign.c(339): error C2133: "d": Unbekannte Größe
Anyone have any idea what could be causing this?
You can use WSL (Windows Subsystem For Linux) to work on a Windows machine while having native Linux. I do that routinely and it works outstandingly well. Native Windows and bioinformatics is just not something that is well-supported.
Ah, that sounds promising. I'll look into this - certainly might be easier than endlessly debugging the install.
Thanks a lot!