Questions about installation

[liu-wq]

Hi Jenny,
I am very interested in this software.I want to reconstruct the ancestral methylation status of the single CpG point.
Unfortunately, I first got stuck during the installation process. Maybe need to adjust the compiled code to match the newer version of c++library. I have made some changes. I don't know if they are correct. I may need you to check if my changes are correct because I am really not good at C++.

1. Although smithlab_cpp and adssrc are linked, these two folders are empty after git clone. I downloaded them separately.
2. error: ‘cbegin’ was not declared in this scope. # I change it to begin. What is the difference between this cbegin? cend also has this issue, I changed it to end.
3. error: ‘copy’ was not declared in this scope. # I change it to std::copy
4. error: ‘SMITHLABException’ was not declared in this scope. # I added "struct SMITHLABException" belonging to repositories "rseg" to smithlab_utils.hpp.
5. error: ‘std::filesystem’ has not been declared # I changed the c++11 in / media/rocket/program/Epiphyte/src/prog/Makefile to c++17. (This is what I am most uncertain about, although it was successfully installed.)

There is also a detail to ask, what is the lowest CpG coverage used in the estimated state? single CpG coverage >= 5?

Best
Weiqiang

[andrewdavidsmith]

First try cloning recursively. If you try to download things separately, the versions will almost certainly mismatch.

[liu-wq]

Thank you!
The software has been installed successfully, but the example file cannot be run successfully. Did I miss a step?

terminate called after throwing an instance of 'std::runtime_error'
  what():  bad methpipe site line: "chr1        19464625        19464626        1111111111111   0.937391        +"
[1]    745700 abort (core dumped)  ../bin/epiphy-sim CpGs.bed sim.params -L sim_leaf.out -o sim_all.out

[liu-wq]

Hi,
I successfully ran the example file. Upon rechecking the code in "MethpipeFiles.cpp", it was found that the CpG.bed given in the first step of methylation simulation was not in the defined input file format, but in the ". meth" format used by Methpipe. I simply converted it to meth format by awk '{OFS="\ t"; print $1, $2, $6, "CpG", $5,100}' CpGs. bed>test. meth, and the first step ran successfully.

The generated "sim_leaf. out" file also solved my confusion about the format of the input file for the "epiphy-est" program. I generated a posterior probability matrix of my own data based on the code provided in the comparison of 7 mammalian sperm HMRs, but it kept running unsuccessfully. After adjusting, I finally started running.

There is also some confusion about whether the posterior probabilities of low-coverage sites (for example, coverage < 5) are credible. Retain or treat as missing values?
Because I don't know much about the calculation principle of this process, I'm not sure what to do is more reasonable.

Best,
Weiqiang