haplotag --log

[bsb2014]

I am wondering if the outputted plain-text file also contains PS tags in addition to HP tags. As you suggested, both pieces of information are important to determine haplotypes. Thanks.

[twolinin]

Hi @bsb2014 ,

We have added the 'PhaseSet' and '(PhaseSet,Variantcount)' to the output of the --log to represent the PS tag to a read.

#Read   Chr     ReadStart       Confidnet(%)    Haplotype       PhaseSet        TotalAllele     HP1Allele       HP2Allele       phasingQuality(PQ)      (Variant,HP)    (PhaseSet,Variantcount)
eb459876-8c81-4714-a496-a90ea8be94d2    chr1    10000   0.969697        1       43042   33      32      1       15       47660,0 47695,0 47698,0 47705,0 47760,0 47929,0 48170,0 48182,0 48936,0 48975,0 49147,0 49242,0 49290,0 49313,0 49314,0 49341,0 49362,0 49403,0 49426,0 49481,0 51402,0 51458,0 51498,0 51619,1 51801,0 51805,0 51901,0 51935,0 51940,0 51950,0 52104,0 52143,0 52524,0         43042,33
6ca3a71f-62fd-416e-8c6e-8c4a9c054e1a    chr1    10000   0.984375        2       43042   64      1       63      18       47660,1 47695,1 47698,1 47705,1 47760,1 47929,1 48170,1 48182,1 48936,1 48975,1 49147,1 49242,1 49290,1 49313,1 49314,1 49341,1 49362,1 49403,1 49426,1 49481,0 51402,1 51458,1 51498,1 51619,1 51801,1 51805,1 51901,1 51935,1 51940,1 51950,1 52104,1 52143,1 52524,1 53008,1 53788,1 53816,1 53957,1 54177,1 54394,1 55387,1 56678,1 57063,1 57681,1 57817,1 57855,1 57989,1 57997,1 58431,1 58447,1 58770,1 58811,1 58865,1 58988,1 59275,1 59343,1 60159,1 60272,1 60331,1 60717,1 60828,1 61218,1 61479,1 61578,1 61735,1         43042,64
0d8b7c68-d98d-4045-a572-82fedac62da5    chr1    10000   -nan    .       .       0       0       0       0

This modification will be released in the upcoming version. For now, you can compile using the following methods.

git clone https://github.com/twolinin/longphase.git
cd longphase
autoreconf -i
./configure
make -j 4

Thanks

[bsb2014]

Many thanks and looking forward to the next binary release.

I can not compile it on the server, and the error is

/usr/bin/ld: cannot find -ldeflate collect2: error: ld returned 1 exit status make: *** [Makefile:57: longphase] Error 1

[bsb2014]

I am wondering if haplotag can support multiple bam files, such as short and long read bams.

If I merged the short and long read bam together for phasing, will the haplotag tag both short reads and long reads to HP:i:1 or HP:i:2?

Thanks

[twolinin]

Hi @bsb2014,

1. To solve the compilation issue, I will attempt to troubleshoot the problem. May I know which operating system you are using?
2. Haplotag involves detecting the overlap between reads and SNPs. In theory, tagging will be performed as long as short reads contain a sufficient number of SNPs.

[bsb2014]

1. I am using Ubuntu 16.04.7 LTS
2. During the phasing, can the short reads bridge the long reads if there are overlaps among them? Thanks

[twolinin]

Hi @bsb2014,

1. LongPhase by default uses zlib, and Linux should have it installed by default. Could you please check if zlib is installed on your system?
2. Theoretically not so much as short reads usually cannot span two SNPs, except for dense SNP regions. However, we haven't tested the results of NGS and TGS co-phasing yet. If you don't mind, could you provide the experimental results for ONT-only versus Illumina/ONT hybrid?

Thanks

[bsb2014]

I will try the ONT-only versus Illumina/ONT hybrid and will let you know the results once done.

[bsb2014]

In my case, the results look very similar.

ONT-only: Heterozygous phased:       85.6%
Illumina/ONT hybrid: Heterozygous phased:       86.5%

May I ask when the next binary will be released? Thanks.

[twolinin]

Hi @bsb2014,

Thank you for sharing the results of Illumina and ONT co-phasing. Based on previous experience, there are many factors that can affect the proportion of phased variants, such as sequencing coverage, sequencing error rate, read length etc. Currently, there is no confirmed binary release date. You can try resolving the compilation issue using a Dockerfile or the following command.

export LD_LIBRARY_PATH=" your libdeflate path containing the file libdeflate.so.0"

[bsb2014]

As for Illumina and ONT co-phasing, how does longphase treat pair-end reads? Does it treat one pair as two seperated reads or as one fragment? Thanks

[twolinin]

Hi @bsb2014

Currently, LongPhase is a software developed based on third-generation sequencing, so the two reads of a pair are treated as separate fragments for processing.

[bsb2014]

I understand that phasing requires a read spanning two variants. I am wondering how many phased variants in a read are required for the haplotag? If a read only has one phased variant, will the read be tagged? Thanks

[ythuang0522]

Yes.