Families with largest size differentials

[Xiaojun928]

Hello,

I'm using CAFE to estimate the gene gain and loss event. I've installed it and it works fine for the example data. However, I encountered the problem when running my own data. Here is the log file:

Command line: cafexp -i iqtree.txt -t iqtree.tre

Filtering families not present at the root from: 5290 to 2929

No root family size distribution specified, using uniform distribution

Optimizer strategy: Nelder-Mead with similarity cutoff
Iterations: 300
Expansion: 2
Reflection: 1

Starting Search for Initial Parameter Values
Lambda: 0.49986067102956
Score (-lnL): inf
Lambda: 0.40553337462359
......
Lambda: 0.5059341117188
Score (-lnL): inf

Families with largest size differentials:
OG0000027: 23
OG0002055: 17
......
OG0000025: 9
OG0000012: 9

You may want to try removing the top few families with the largest difference
between the max and min counts and then re-run the analysis.

Failed to initialize any reasonable values

I tried to remove orthologous groups with differential >10. However, the error still exists.

The gene count file was created using orthofinder. Do you have any idea how to deal with such problem?

Thanks~
Xiaoyuan

[evolbeginner]

Are you analyzing gene families of bacterial genomes? I might be wrong but it seems to me that cafe may be more suitable for analyzing those with many duplicates, which may not be common for bacteria.

[Xiaojun928]

Are you analyzing gene families of bacterial genomes? I might be wrong but it seems to me that cafe may be more suitable for analyzing those with many duplicates, which may not be common for bacteria.

Hi,

Thank you for your reply. I'm using CAFE for bacterial genomes. Maybe I have to choose other softwares.

Many thanks.
Xiaoyuan

[benfulton]

It's not wrong to analyze bacteria, but it may not be as accurate. Dr. Hahn's main concern would be that lots of gene gain in bacteria is via horizontal gene transfer, not duplication - CAFE doesn't model HGT. Also, bacteria trees are difficult to work with because they can be extremely deep in time, and difficult to accurately date as the fossils all look the same.

[caord-1]

Hello, I also encountered this problem. I want to know how you have solved it. please!

[Wu-tz]

Hello，have you solved this problem? I used animal genomes but still no solved it when I removed families with largest size differentials.

[benfulton]

You might look at some of the suggestions in #132