lkytal / predfull Goto Github PK

Hello,

I am encountering the following error while running PredFull with tensorflow 2.3.0:

ValueError: Unknown layer: SlicingOpLambda

Is there anyway you could share your conda environment so I can run the model?

Thanks,
Kevin

Hello,

Some of my input peptides which are shorter than length 30 were ignored during prediction. For example, KPIQRTILM(O)M(O)GRYVEPIEDVPCGNI was ignored because the M(O) was interpreted as 4 instead of 1 amino acids long.

I did a quick fix by replacing line 200 in predfull.py with the following, but there is probably a cleaner way to do this
pep_len = 0
for char in item.Peptide:
if char in Alist:
pep_len += 1
#if len(item.Peptide) > max_len:
if pep_len > max_len:

Thank you for this great tool,
Kevin

Hi, I want predicting high resolution ms2 of all tetrapeptides. But now the output result is low resolution using the pm.h5 model . Could you tell me how to get the high resolution ms2 spetrum.

Thanks!

您好
我是爱丁堡大学人工智能专业今年的硕士毕业生。
我今年的毕业设计项目和您的这个项目很相似。我想复现一下您的CNN的结构，所以我能不能看看您的模型的构建代码？
我的邮箱：[email protected] ，如果可以的话非常感谢~

Hello,

I saw your notes said PredFull supports predicting nontryptic peptides. Does this mean you retrained the model with nontryptic peptides? Or just that it is possible to provide them as input?

Thanks,
Kevin

Could you add an example to show how to train a model from scratch?

Bo