fix type incompatible bug when nochr is used

[RunpengLuo]

combin_counts.py always use str type for #CHR, as shown below.

hatchet/src/hatchet/utils/combine_counts.py

Lines 110 to 112 in 3729a35

	bbs = [pd.read_table(bb, dtype={"#CHR": str}) for bb in outfiles]
	big_bb = pd.concat(bbs)
	big_bb = big_bb.sort_values(by=["#CHR", "START", "SAMPLE"])

When nochr is enabled (reference/bam file doesn't have chr prefix), to_dataframe() in rd_gccorrect.py will automatically use int64 dtype for column #CHR, which is incompatible with #CHR from dataframe bb, as shown below and from the issue #236.

hatchet/src/hatchet/utils/rd_gccorrect.py

Lines 20 to 32 in 3729a35

	bb = bb.merge(
	BedTool.from_dataframe(bb[["#CHR", "START", "END"]].drop_duplicates())
	.nucleotide_content(fi=ref_genome)
	.to_dataframe(disable_auto_names=True)
	.rename(
	columns={
	"#1_usercol": "#CHR",
	"2_usercol": "START",
	"3_usercol": "END",
	"5_pct_gc": "GC",
	}
	)[["#CHR", "START", "END", "GC"]]
	)

This fix converts the dtype of #CHR to str post to_dataframe().