The output of ananse network is wried

[linfanxiao]

Dear developers,

I encountered a problem like this: The output of ananse network is wried

head -n 5 Di.network.txt 

tf_target   prob
AHR—A1BG    0.492401
AHR—A1BG-AS1    0.29674652
AHR—A2M 0.3372815
AHR—A2M-AS1 0.25697538

2024-11-19 20:50:35 | INFO | Loading expression data
2024-11-19 20:50:35 | INFO | 73% of TFs found in both BED and expression file(s)
2024-11-19 20:50:35 | DEBUG | Loading tf binding activity data
2024-11-19 20:50:35 | INFO | Loading binding data
2024-11-19 20:50:35 | INFO | Using all 828 TFs.
2024-11-19 20:50:38 | INFO | Using all 1268775 regions.
Aggregating on chr17. Overall progress: 100%|██████████| 24/24 [25:16<00:00, 63.17s/contig]
2024-11-19 21:17:59 | INFO | Processing expression-binding network
2024-11-19 21:18:55 | DEBUG | Using tf_expression, target_expression, weighted_binding, activity
2024-11-19 21:18:56 | INFO | Writing network

my command is like this:

ananse network -e Di*.txt -n 1 -o Di.network.txt Di.binding/binding.h5

https://github.com/vanheeringen-lab/ANANSE/issues/116 I found someone has a similar question. However, they don't show how to solve it.

[linfanxiao]

I know what makes this error: """ Infer cell type-specific gene regulatory network.

Parameters

genome : str The genome used for gene annotation and enhancer location. (default: "hg38") gene_bed : str, optional Gene annotation for the specified genome as a 12-column BED file. (default: None) include_promoter : bool Include or exclude promoter peaks (<= TSS +/- 2kb) in network inference. (default: False) include_enhancer : bool Include or exclude enhancer peaks (> TSS +/- 2kb) in network inference. (default: True) full_output : bool Export all variables to the GRN file, or by default, only the TF_target + prob score.

Note

Using the f parameter should be specified either in the network and inference steps or in neither. """