hk_parse_seg error

[bredeson]

Hi!,

I'm attempting to convert a .pairs file to a .3gd file but am recieving the following error:

hickit: io.c:81: hk_parse_seg: Assertion `c != 0' failed.
./infer-3dgc.sh: line 5: 22408 Aborted                 ./hickit-0.1_x64-linux/hickit -i mnd.pairs.gz -r1m -c1 -r10m -c5 -b4m -b1m -b200k -D5 -b50k -D5 -b20k -O mnd+flt.3dg

I'm using the binaries provided here that I downloaded today (May 19, 2019). My pairs file looks like:

.	Chr1	1	.	Chr1	178	+	-	1	0	0	0
.	Chr1	7	.	Chr1	236	+	-	1	0	0	0
.	Chr1	11	.	Chr1	212	+	-	1	0	0	0
.	Chr1	11	.	Chr1	272	+	-	1	0	0	0
.	Chr1	25	.	Chr1	113	+	-	1	0	0	0
.	Chr1	35	.	Chr1	271	+	-	1	0	0	0
.	Chr1	60	.	Chr1	422	+	-	1	0	0	0
.	Chr1	61	.	Chr1	218	+	-	1	0	0	0
.	Chr1	64	.	Chr1	266	+	-	1	0	0	0
.	Chr1	73	.	Chr1	169	+	-	1	0	0	0
.	Chr1	79	.	Chr1	355	+	-	1	0	0	0
.	Chr1	82	.	Chr1	292	+	-	1	0	0	0
.	Chr1	103	.	Chr1	303	+	-	1	0	0	0
.	Chr1	108	.	Chr1	284	+	-	1	0	0	0
.	Chr1	111	.	Chr1	230	+	-	1	0	0	0
.	Chr1	122	.	Chr1	262	+	-	1	0	0	0
.	Chr1	130	.	Chr1	167	+	-	1	0	0	0
.	Chr1	141	.	Chr1	501	+	-	1	0	0	0
.	Chr1	142	.	Chr1	284	+	-	1	0	0	0
.	Chr1	142	.	Chr1	440	+	-	1	0	0	0

(The genome is 17 generations inbred, and so I faked the phase probabilities above). I tried running hickit using a pairs file with and without the probabilities, but receive the same error each time.

Any help you can offer me in getting to the next step would be very much appreciated.

Thank you!

[lh3]

You have to have header.

[bredeson]

Your recommendation helped me get past my previous error, thank you. Now, however, I am receiving the following error:

[M::hk_map_read] read 218057034 pairs
[W::hk_sd_ploidy_XY] no sex chromosomes (identified by 'X' or 'Y' in chr names)
[M::hk_pair_filter_close_legs] filtered out 225387/218057034 pairs with close legs
[M::hk_pair_dedup] duplicate rate: 87.67% = 190983876 / 217831647
[M::hk_pair_filter_isolated] threshold: 1; filtered out 4809/26847771 isolated pairs
[M::hk_pair_filter_isolated] threshold: 5; filtered out 1269/26842962 isolated pairs
hickit: bin.c:34: hk_bmap_set_offcnt: Assertion `m && m->d && m->beads' failed.
./infer-3dgc.sh: line 5: 31004 Aborted                 ./hickit-0.1_x64-linux/hickit -i mnd.pairs.gz -r1m -c1 -r10m -c5 -b4m -b1m -b200k -D5 -O mnd+flt.3dg

Does this mean my data are too sparse? Or have I set my flags incorrectly?

I very much appreciate your feedback.

[tanlongzhi]

Thank you for your interest in our work. Would you like to attach your pairs file, so we can troubleshoot?

[cyz0315]

I received the same problem, as hickit: bin.c:34: hk_bmap_set_offcnt: Assertion `m && m->d && m->beads' failed.
Pairs file is attached.
GSM3530987_181102_snm3C_A1_AD002_indexed_contacts.txt

[tanlongzhi]

@cyz0315 This file is not in pairs format either. See this file from our latest publication for example.

[cyz0315]

Sorry, I uploaded wrong file. There was a problem uploading .pairs file, so I changed to .txt file.
GSM3530987_181102_snm3C_A1_AD002.pairs.txt

[lh3]

See above:

You have to have header.

[lh3]

Also, you have to have phasing information and follow the pipeline in README.

[tanlongzhi]

Also, you have to have phasing information

unless the sample is haploid, which still doesn't apply given the GEO accession you're referring to (they used a diploid mouse cell line).

By the way, this repo cannot be directly applied to that dataset (sn-m3C-seq), because during their bisulfite chemistry, most Cs are converted to Ts in a strand-specific manner--thus interfering with our haplotype determination. This will certainly be a useful functionality to add in the future; but we don't support it yet.