No valid result obtained for this network

[UhP88]

Thank you for this helpful package and tutorial... I am trying to construct a network analysis and plot with below steps. However I get an error on the way. Below is the code and warnings/error messages I get. What could be the reason for not getting valid results.

Thank you in advance

dataset <- microtable$new(sample_table = MicroEco_SampleInfo, otu_table = MicroEco_Species_Count) tY <- trans_network$new(dataset = dataset, cor_method = "spearman", filter_thres = 0.001) tY$cal_network(COR_p_thres = 0.01, COR_cut = 0.6) ---------------- 2023-07-18 09:40:16 : Start ---------------- Perform p value adjustment with fdr method ... ---------------- 2023-07-18 09:40:16 : Finish ---------------- The result network is stored in object$res_network ... Warning message: In length(vattrs[[name]]) <- vc : length of NULL cannot be changed tY$cal_module() Totally, 3 modules are idenfified ... Modules are assigned in network with attribute name -- module ... tY$get_node_table() The nodes (6) with NaN in z will be filtered ... Error in private$assign_module_roles(zp) : No valid result obtained for this network! tY$plot_network(method = "ggraph") Run get_node_table function to get or update the node property table ... The nodes (6) with NaN in z will be filtered ... Error in private$assign_module_roles(zp) : No valid result obtained for this network!

[ChiLiubio]

Hi. Could you please attach your dataset so that I can reproduce your result? Please save your dataset following the tutorial (https://chiliubio.github.io/microeco_tutorial/notes.html#save-function). Please also provide your microeco version. Thanks.

[UhP88]

Thank you so much for your prompt reply. Below attached is the test data file and the micro version is 0.8.0

Thank you again...

On Jul 18, 2023, at 10:59, Chi Liu @.***> wrote:

Hi. Could you please attach your dataset so that I can reproduce your result? Please save your dataset following the tutorial (https://chiliubio.github.io/microeco_tutorial/notes.html#save-function https://chiliubio.github.io/microeco_tutorial/notes.html#save-function). Please also provide your microeco version. Thanks.

— Reply to this email directly, view it on GitHub https://github.com/ChiLiubio/microeco/issues/259#issuecomment-1639163096, or unsubscribe https://github.com/notifications/unsubscribe-auth/ASTKPYUWVTCCUFAF2HJWHSTXQXUX3ANCNFSM6AAAAAA2NU4FWQ. You are receiving this because you authored the thread.

[ChiLiubio]

Hi. I find nothing in the attachment. Please attach the compressed data in this github issue page. Thanks.

[UhP88]

Sorry for the error.. The data set was not attached previously as I replied through email. dataset_TEST.zip

[ChiLiubio]

Hi. Please first update your microeco package to v0.20.0 and try again. I have tested your steps with my package v0.20.0. All the steps are ok. Please tell me if the error is still there.

[UhP88]

I updated the package and did the analysis again.. I am getting the same error when using gggraph. But I was able to save the results as an igraph and view through Gephi.. Thank you for your help!!!

[finally-jay]

Hi Chi, I am now using the v1.0.0, and got the similar error when running this line : t1$get_node_table(node_roles = TRUE)

And my network only contained 2 nodes and 1 edge, may I ask for some advice, is it because of my data or the way I preprocessed it (I rarefied to even depth) ?

Here's my code:

> ps <- phy_1wk
> sample_data(ps)$Group <- sample_data(ps)$cs_1y
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
> sample_data(ps)$Group <- relevel(sample_data(ps)$Group, ref = "Low")
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
Found more than one class "phylo" in cache; using the first, from namespace 'phyloseq'
Also defined by ‘tidytree’
> otu <- otu_table(ps)
> dim(otu)
[1] 25975   498
> summary(rowSums(otu) == 0)
   Mode   FALSE    TRUE 
logical    1895   24080 
> 
> meco_dataset <- phyloseq2meco(ps)
24080 taxa with 0 abundance are removed from the otu_table ...
> meco_dataset$tax_table %<>% .[grepl("Bacteria|Archaea", .$domain), ]
> meco_dataset$tax_table$Phylum <- meco_dataset$tax_table$phylum
> 
> meco_dataset$tidy_dataset()
> print(meco_dataset)
microtable-class object:
sample_table have 498 rows and 71 columns
otu_table have 1865 rows and 498 columns
tax_table have 1865 rows and 11 columns
phylo_tree have 1865 tips
rep_fasta have 1865 sequences
> 
> meco_dataset$sample_sums() %>% range #2020 197004  -> 2020 2020
[1]   2020 197004
> 
> meco_dataset$rarefy_samples(sample.size = 2020)
652 features are removed because they are no longer present in any sample after random subsampling ...
652 taxa with 0 abundance are removed from the otu_table ...
> # The parameter cor_method in trans_network is used to select correlation calculation method.
> # default pearson or spearman correlation invoke R base cor.test, a little slow
> t1 <- trans_network$new(dataset = meco_dataset, cor_method = "spearman", filter_thres = 0.05)
After filtering, 4 features are remained ...
The correlation result list is stored in object$res_cor_p ...
> # return t1$res_cor_p list, containing two tables: correlation coefficient table and p value table
> 
> t1 <- trans_network$new(dataset = meco_dataset, cor_method = "spearman", use_WGCNA_pearson_spearman = TRUE, filter_thres = 0.05)
After filtering, 4 features are remained ...
The correlation result list is stored in object$res_cor_p ...
> 
> # construct network; require igraph package
> t1$cal_network(COR_p_thres = 0.05, COR_optimization = TRUE)
---------------- 2023-08-02 16:52:45.777463 : Start ----------------
Perform p value adjustment with fdr method ...
Start COR optimizing ...
[1] 0.05
[1] 0.1
[1] 0.2
[1] 0.25
[1] 0.3
[1] 0.35
[1] 0.4
[1] 0.45
[1] 0.5
[1] 0.6
[1] 0.65
[1] 0.7
[1] 0.75
[1] 0.8
The optimized COR threshold: 0.05...

---------------- 2023-08-02 16:52:45.872099 : Finish ----------------
The result network is stored in object$res_network ...
> # use arbitrary coefficient threshold to contruct network
> # t1$cal_network(COR_p_thres = 0.01, COR_cut = 0.7)
> # return t1$res_network
> 
> 
> # invoke igraph cluster_fast_greedy function for this undirected network 
> t1$cal_module(method = "cluster_fast_greedy")
Use cluster_fast_greedy function to partition modules ...
Totally, 2 modules are idenfified ...
Modules are assigned in network with attribute name -- module ...
> 
> library(rgexf)
> # require rgexf package to be installed
> t1$save_network(filepath = "network.gexf")
> 
> # calculate network attributes
> t1$cal_network_attr()
Result is stored in object$res_network_attr ...
> t1$res_network_attr
> 
> # get node properties
> t1$get_node_table(node_roles = TRUE)
Error in private$assign_module_roles(zp) : 
  No valid result obtained for this network!



Thank you for you time !!
Jay

[xinguo1236]

I am now using the v1.4.0, and got the similar error when running this line : soil_amp_network %<>% get_node_table(node_roles = TRUE) %>% get_edge_table， Error in private$assign_module_roles(zp) : No valid result obtained for this network! Thank you for you time !!