How to run flame by creating a singularity sif file

[zillurbmb51]

Hi,

I wanted to use this nice tool for one of my projects. I can not use docker/git clone because my hpc can not fetch anything from online. So, I have created a flame.sif file from my laptop and transfer this sif to my hpc. singularity pull flame.sif docker://pavlopouloslab/flame

Now, how can I run it?

singularity exec /path/to/flame.sif then what? Any instruction for me? (I have done this for many other tools which has executables, but this time I am not sure what to do)

I tried it from my laptop following the instructions using git clone. It leads me to a GUI. Is there any command line options. I have many lists of snps for multiple traits. It is impossible to paste snps for each trait.

Best, Zillur

[fbaltoumas]

Hello,

Thank you for taking an interest in Flame.

Unfortunately, it looks like the tool will not be able to work for you in the way you describe for two reasons:

1. Right now, Flame is a UI-only tool. We have plans for a package-like version, which users will be able to import into their scripts/workflows, but this is something to be implemented for a future update and does not exist for the current version.

2. You say that your HPC "cannot fetch anything from online". If this means that your system does not have any online capabilities to send/receive requests, then that means that Flame won't work on it either. That's because the tool uses a number of API requests to online services (Ensembl, Gene Ontology, dbSNP, aGOtool etc) as part of its main functionality.

Having said that, and depending on the number of SNPs / SNP datasets you have, you could consider using the API of the online version. Through that, you can automatically upload multiple sets of lists to Flame programmatically.

You can use POST requests to upload your lists, formatted as a JSON object like the one below:

{
  "gene_list1": ["GNAS", "ABCG2", "WT1", "CDK2", "FLT1", "CCN2", "INSR"],
  "gene_list2": ["MMP1", "PTGS2", "C3", "PON1", "LDLR", "HBA1", "CYP1B1", "PTEN", "SNCA", "RAC1", "BCL2", "HLA-DRB1", "IL13", "GRN", "NF1", "AHR", "YAP1", "LCN2", "FOXO3", "LEPR", "ABCB1", "STAT3", "HRAS", "NOS3", "HLA-DQA1", "BRCA1", "EZH2", "CDKN1B"],
  "gene_list3": ["HLA-C", "PTEN", "UGT1A1", "CDH1", "MDM2", "EGFR", "FMR1", "VEGFA", "ERCC1", "VWF", "CFH", "OGG1", "TGFB1", "IGF2", "ADRB2", "AGER", "CETP", "CYP2D6", "KIT", "HTT", "ACE2", "CASP8", "HDAC1", "GJB2", "IL4", "EDN1", "OPRM1", "NOS3", "ERCC2", "NFKB1", "KCNQ1"]
}

(In this example, we have genes, but you can do the same with SNPs).

The API endpoint link is this one:

https://pavlopoulos-lab-services.org/bib/api/flame

Afterwards, you can either convert the SNP lists to genes through the "Conversion" -> "Gene ID Conversion" function from the sidebar menu, or you can run directly enrichment to the SNPs themselves.

[zillurbmb51]

@fbaltoumas

Thank you so much.

I tried the API version. I have a json file of snps more than a thousand phenotypes.

# Define the path to your JSON file
json_file_path <- "../../adhd_metabolomics/snp2function/adhd_metabo_conjfdr_snplist.json"

# Read the JSON file into R
data_to_send <- fromJSON(json_file_path)

# Print the JSON data (optional, for verification)
cat("JSON to send:\n", toJSON(data_to_send, pretty = TRUE), "\n")

# Send a POST request
response <- POST("https://pavlopoulos-lab-services.org/bib/api/flame",
                 body = data_to_send,
                 encode = "json",
                 content_type_json())

# Check the response status
if (http_status(response)$category == "Success") {
  print("Data uploaded successfully.")
} else {
  print(paste("Failed to upload data:", http_status(response)$message))
}

# Print the content of the response
print(content(response, "text"))

I got the api link and when I opened the link on the browser it opens a same page as the web version with my lists but it only takes 10 lists and for conversion I can do one list at a time.

How can I automate the process or submit altogether?

Best, Zillur