运行plotCoverage.py遇pysam模块缺失，安装pysam时出现os/rand.c文件不存在错误的解决求助

I’ve run into this exact compilation issue with pysam before, and it’s usually tied to missing system dependencies or quirks in how pip handles htslib builds under the hood. Here are a few tried-and-true fixes to get you past this:

1. Install all required system dependencies for htslib #

liblzma-dev is one piece of the puzzle, but htslib (the core library pysam relies on) needs several other dependencies to compile successfully. Run this command to install the full set:

sudo apt-get update && sudo apt-get install libbz2-dev libcurl4-openssl-dev libssl-dev zlib1g-dev python3.9-dev

• python3.9-dev ensures critical Python header files (like Python.h) are available, which also addresses your earlier deeptoolsintervals error.
• The other libraries handle compression (bz2, zlib, lzma) and network/crypto functionality htslib depends on.

2. Use Conda to install pysam (avoids compilation entirely) #

If compiling from source keeps failing, using Conda’s pre-built pysam package is often the easiest workaround. Conda handles all dependency conflicts and delivers pre-compiled packages tailored to your system:

# If you don’t have Conda set up, install Miniconda first, then activate your environment:
conda activate your_working_env
conda install -c bioconda pysam

Bioconda is the standard channel for bioinformatics tools like pysam, so this should pull in a fully functional version without any build errors.

3. Force pip to use system-wide htslib (bypass isolated builds) #

If you must use pip, try disabling its build isolation feature—this makes pip use your system’s pre-installed htslib instead of trying to compile it from scratch. First, install htslib system-wide:

sudo apt-get install htslib-dev

Then install pysam with this flag:

pip3.9 install pysam --no-build-isolation

This tells pip to skip downloading and building its own copy of htslib, relying on the system’s pre-configured version instead.

4. Verify your Python environment alignment #

Double-check that you’re installing pysam into the same Python 3.9 environment you’re using to run plotCoverage.py. It’s easy to accidentally install packages into a different Python version without noticing. Run these commands to confirm alignment:

which python3.9
which pip3.9

If the paths don’t match, use python3.9 -m pip install pysam instead of pip3.9—this guarantees you’re targeting the correct Python instance.

内容的提问来源于stack exchange，提问作者xyz0o