Inconsistency in results about nichenetr HOT 9 CLOSED

Hi @aaronkwc

Thank you for pointing this out. This is indeed concerning.
I have never come across a problem like this, so I would need more information about where exactly in the NicheNet analysis pipeline differences start to occur.

Because the end result shows a different ligand ranking this has probably to do with one of the two following:

1. differences in potentially active ligands
2. differences in the geneset of interest and/or background

So I suggest that you first check if you observe differences between the machines in one of these.
This is rather easy to check if you used the step-by-step analysis. If you ran the seurat wrapper, you can find all ligands via nichenet_output$ligand_activities$test_ligand, however it is now not directly possible to extract the geneset of interest and the background from the output (I will change this later today, but then you will need to reinstall nichenetr)

Looking forward to your answer.

Edit: If you reinstall nichenetr now, and run the analysis again with the seurat wrapper, you will be able to directly extract the geneset of interest

from nichenetr.

Thanks for the reply and update @browaeysrobin

Here are the test ligands:

A$ligand_activities$test_ligand
[1] "F9" "Nucb2" "Serpine1" "Serpinc1" "Slit2" "Pdgfd"
[7] "Nptn" "Psap" "Nectin3" "Inha" "Nectin2" "Timp1"
[13] "Ttr" "Angptl3" "Pcdhb22" "Sct" "Nampt" "Trf"
[19] "Alcam" "Celsr2" "Ptdss1" "Fgf20" "Manf" "Apoa1"
[25] "Tnfsf13" "Inhbe" "Gdf15" "Adam23" "Cxcl2" "Serpine2"
[31] "Adm2" "Nppc" "Gdf9" "Sema7a" "Grn" "Dsc1"
[37] "Vegfd" "Sema3c" "Sema4d" "Chgb" "Rarres2" "Adm"
[43] "Sema3f" "Ptprf" "Nlgn2" "Nrxn1" "C3" "Lamc3"
[49] "Proc" "Hebp1" "Nfasc" "Serping1" "Cdh2" "Vtn"
[55] "F11r" "Rabep1" "Arf1" "Reln" "Npb" "Ptgs2"
[61] "Madcam1" "Npff" "Pcdh7" "Nmb" "Serpina1a" "Rtn4"
[67] "Ccl25" "Aimp1" "Mst1" "Sema3e" "Nenf" "F7"
[73] "Fgf9" "Cxcl16" "Cntn2" "Rabep2" "Flrt3" "Nts"
[79] "F8" "Cadm3" "Icam5" "Npw" "Tff1" "Clec11a"
[85] "Il16" "Eda" "Apob" "Cxcl13" "Lgals3" "Amigo1"
[91] "Fgf21" "Nectin1" "Fgf11" "H2-T23" "Artn" "Uts2b"
[97] "Flrt2" "Agrn" "Efna5" "Angptl1" "Pvr" "Plxnb2"
[103] "Ren1" "Dusp18" "Robo1" "Icam2" "Mapt" "Fgf12"
[109] "Cd320" "Rspo3" "Sdk1" "Hspg2" "Hbegf" "Hsp90b1"
[115] "Cxcl5" "Ntn4" "Lamb1" "Amigo2" "Uts2" "Cdhr5"
[121] "Ccl28" "Plau" "Dsc2" "Cntnap1" "Pf4" "Liph"
[127] "Saa1" "Clstn1" "Itgb2" "Ccl5" "Angpt2" "Celsr1"
[133] "Pdgfc" "Cmtm4" "Lif" "Ccl20" "Ereg" "Cmtm6"
[139] "Lamc2" "Yars" "Wnt5a" "L1cam" "Wnt9b" "Wisp3"
[145] "a" "Cyr61" "Mdk" "Gstp2" "Ccl19" "Itga9"
[151] "Jam3" "Ccl4" "Vcam1" "Cdh1" "Nrg2" "Fga"
[157] "Fgf7" "Dsg2" "Plekho2" "Nrg4" "Gal" "Nid1"
[163] "Nrg1" "Gnrh1" "Fbn1" "Kitl" "Efnb1" "Psen1"
[169] "Ceacam1" "Lrpap1" "Pdgfa" "Efnb2" "Areg" "Alox5ap"
[175] "Xcl1" "Ccl24" "Sell" "Vegfc" "Cga" "Bmp6"
[181] "Ccl2" "Gpi1" "Pik3cb" "Dchs1" "Col18a1" "Il34"
[187] "Nrtn" "Cadm4" "Adam15" "Ctsg" "Ctgf" "Plat"
[193] "Icosl" "Hdc" "Efna1" "Oit1" "Prok1" "Bmp7"
[199] "Slit1" "Pecam1" "Ncam1" "Cx3cl1" "Efna3" "Vegfb"
[205] "Cxcl9" "Cxcl10" "Efna4" "Dsg1a" "Gmfb" "Spn"
[211] "Efnb3" "Calm2" "Hgf" "Tnfsf9" "Ptprc" "Icam1"
[217] "Cdhr2" "Rps19" "Adam12" "Wnt9a" "Bmp4" "Rgma"
[223] "Epo" "Pkd1" "Tnfsf10" "Ppbp" "Itgb1" "Cdh5"
[229] "Cdh3" "Ptprm" "Selplg" "Thbs1" "A2m" "Sema5a"
[235] "Tnfsf13b" "Il1f9" "Efna2" "Txlna" "Icam4" "Hc"
[241] "Dlk1" "Kng1" "Btc" "Pros1" "Plxnb3" "Il11"
[247] "Anxa1" "Il1rn" "Ntf5" "Camp" "Fbrs" "Sele"
[253] "Calm1" "Sema3b" "Col1a1" "Efemp1" "Cxcl1" "Csf1"
[259] "Ntf3" "Pomc" "Cfh" "Thbs2" "Insl6" "Selp"
[265] "Shh" "Adam9" "Tnc" "Apoe" "Inhbc" "Ccl7"
[271] "Pcdh1" "Ccl1" "Vwf" "Tgfa" "Ngf" "Gas6"
[277] "Ret" "Hamp" "Copa" "Itgb7" "Il7" "Cxcl12"
[283] "Il33" "Ostn" "Vegfa" "Col4a1" "Flt3l" "Fn1"
[289] "Hdgf" "Tfpi" "Bmp1" "Pdgfb" "Plg" "Col5a3"
[295] "Sema3a" "Jam2" "Cxcl14" "Esam" "Fat1" "Fgf1"
[301] "Angpt1" "Tnfsf12" "Inhba" "Bmp2" "Adam17" "Lamb2"
[307] "Cd200" "Sema6d" "Hmgb1" "Ocln" "Bdnf" "Glg1"
[313] "Tgfb3" "Crtam" "F2" "Fgf23" "Mag" "Spp1"
[319] "Edn1" "Ltb" "Il1a" "Il1b" "Inhbb" "Cdh6"
[325] "Il18" "Dll1" "Clcf1" "Cadm1" "Epgn" "Pcdhgc3"
[331] "Il4" "Tnf" "Dll4" "Crlf1" "Nodal" "Csf3"
[337] "Osgin1" "Lefty1" "Tgfb1" "Tnfsf14" "Bmp5" "Il23a"
[343] "Gdf3" "Cmtm8" "Sfrp2" "Ihh" "Itgal" "Ptprt"
[349] "Wisp2" "Gpnmb" "Ebi3" "Ctf1" "Tslp" "Dll3"
[355] "Igf1" "Crlf2" "Il19" "Agt" "Jag1" "Hmgb2"
[361] "Egf" "Il15" "Cd24a" "Lama2" "Tgfb2" "Mpz"
[367] "Apln" "Ltf" "Calr" "Mif" "App"

B$ligand_activities$test_ligand
[1] "Cd6" "F9" "Serpinc1" "Serpine1" "Nucb2" "Nectin3"
[7] "Nampt" "Nptn" "Angptl3" "Psap" "Inha" "Pdgfd"
[13] "Timp1" "Nectin2" "Pcdhb22" "Slit2" "Ttr" "Trf"
[19] "Alcam" "Celsr2" "Apoa1" "Tnfsf13" "Inhbe" "Adm"
[25] "Sema4d" "Vegfd" "Adm2" "Sema7a" "Grn" "Manf"
[31] "Cxcl2" "Nlgn2" "Nrxn1" "Gdf9" "Vtn" "Hebp1"
[37] "Dsc1" "Adam23" "Sema3f" "Serping1" "Rarres2" "Ptprf"
[43] "Ptdss1" "C3" "Wnt4" "Ccl25" "Ptgs2" "Aimp1"
[49] "Proc" "Arf1" "Chgb" "Nfasc" "Npff" "Npb"
[55] "Reln" "F11r" "Nmb" "Lamc3" "Lgals3" "Bmp15"
[61] "Cdh2" "Cxcl16" "Rabep1" "Cadm3" "Pcdh7" "Sema3e"
[67] "F8" "Madcam1" "Nectin1" "F7" "Penk" "Rabep2"
[73] "Cxcl13" "Plxnb2" "Wnt7b" "Fgf21" "Hbegf" "Nenf"
[79] "Mst1" "Amigo1" "Icam5" "Serpina1a" "Apob" "Saa1"
[85] "H2-T23" "Pvr" "Efna5" "Clec11a" "Fgf7" "Cd320"
[91] "Prok1" "Artn" "Cxcl5" "Yars" "Hspg2" "Podxl2"
[97] "Hsp90b1" "Agrn" "Celsr1" "Fga" "Cntnap1" "Amigo2"
[103] "Dsc2" "Ntn4" "Dscam" "Lif" "Angpt2" "Robo1"
[109] "Cmtm4" "Ccl20" "Nrg2" "Vcam1" "Kitl" "L1cam"
[115] "Mapt" "Wisp3" "Ccl4" "Cmtm6" "Gal" "Liph"
[121] "Wnt5a" "Sdk2" "Nrg4" "Jam3" "Ccl19" "Uts2"
[127] "Rtn4" "Gstp2" "Icam2" "Lamc2" "Plau" "Dsg2"
[133] "Negr1" "Efnb1" "H2-M3" "Ceacam1" "Ccl2" "Pdgfc"
[139] "Hc" "Cdhr5" "Efnb2" "Cxcl10" "Gnrh1" "Icam1"
[145] "Plekho2" "Pcdhga11" "Mdk" "Clstn1" "Cdh1" "Lama1"
[151] "Sell" "Psen1" "Vegfc" "Il34" "Nrtn" "Cyr61"
[157] "Cxcl9" "Pik3cb" "Fat4" "Lrpap1" "Nid1" "Areg"
[163] "Pdgfa" "Plat" "Amh" "Gpi1" "Col18a1" "Cadm4"
[169] "Ctgf" "Oit1" "Efna1" "Dsg1a" "Icosl" "Ucn2"
[175] "Cga" "Ctsg" "Efna3" "Bmp7" "Ncam1" "Hdc"
[181] "Efna4" "Vegfb" "Nrg1" "Bmp4" "Cx3cl1" "Gmfb"
[187] "Anxa1" "Adam15" "Slit1" "Sema5a" "Wnt9a" "Rps19"
[193] "Btc" "Rgma" "Pkd1" "Calm2" "Selplg" "Cdh5"
[199] "A2m" "Hgf" "Apoe" "Il11" "Itgb1" "Il1rn"
[205] "Epo" "Tnfsf13b" "Efna2" "Ntf5" "Cd274" "Ccl11"
[211] "Txlna" "Icam4" "Csf1" "Tnfsf10" "Il1f9" "Pros1"
[217] "Thbs1" "Dlk1" "Kng1" "Sema3b" "Thbs2" "Ptprm"
[223] "Calm1" "Insl6" "Cxcl1" "Tnc" "Col1a1" "Ntf3"
[229] "Selp" "Il33" "Fbrs" "Shh" "Fasl" "Inhbc"
[235] "Cdh15" "Cxcl12" "Cfh" "Cdhr2" "Pcdh1" "Adam9"
[241] "Osm" "Gas6" "Ccl7" "Vwf" "Pomc" "Tgfa"
[247] "Col4a1" "Vegfa" "Hamp" "Copa" "Adam17" "Ret"
[253] "Ngf" "Col5a3" "Lep" "Fat1" "Plg" "Fn1"
[259] "Tfpi" "Inhba" "Hmgb1" "Jam2" "Tnfsf12" "Hdgf"
[265] "Ocln" "Cd200" "Fgf1" "Lamb2" "Bmp2" "Glg1"
[271] "Bdnf" "F2" "Angpt1" "Tgfb3" "Crtam" "Spp1"
[277] "Edn1" "Il1a" "Mag" "Inhbb" "Ltb" "Il18"
[283] "Clcf1" "Ifnk" "Epgn" "Dll1" "Cadm1" "Ifnb1"
[289] "Pcdhgc3" "Pcdhga7" "Dll4" "Nodal" "Crlf1" "Osgin1"
[295] "Csf3" "Tgfb1" "Tnfsf14" "Lefty1" "Bmp8a" "Bmp5"
[301] "Il23a" "Sfrp2" "Cmtm8" "Ihh" "Itgal" "Ptprt"
[307] "Gpnmb" "Ebi3" "Tslp" "Ctf1" "Agt" "Igf1"
[313] "Wisp2" "Jag1" "Il15" "Egf" "Cd24a" "Hmgb2"
[319] "Tgfb2" "Lama2" "Mpz" "Calr" "Mif" "App"

Also, I tried re-installing Nichenet on my desktop, but it gave me errors:

error: failed to install 'nichenetr' from github: missing value where true/false needed

I am able to reinstall it on my laptop, but it does not have enough RAM to reran the analysis....any idea on how to fix it? My suspicion is it has something to do with Rstudio, but I'm not sure.

from nichenetr.

Hi @aaronkwc

I would suggest deleting and reinstalling Rstudio and trying to reinstall nichenetr from github again. If it does not get installed from Rstudio, try installing it from the R console. Can you also check which R versions you use? And how many RAM you have?

from nichenetr.

Thanks @browaeysrobin

Sorry for the delayed reply, I have tried most available solutions found on Google, including

1. Reinstall Rstudio
2. Reinstalling R
3. Installing from R console
4. Reinstalling devtools

It still gives me the same error:
`- preparing 'nichenetr':
checking DESCRIPTION meta-information ...

checking DESCRIPTION meta-information ...

√ checking DESCRIPTION meta-information

• checking for LF line-endings in source and make files and shell scripts

• checking for empty or unneeded directories

• looking to see if a 'data/datalist' file should be added

• building 'nichenetr_0.1.0.tar.gz' (2.2s)

Error: Failed to install 'nichenetr' from GitHub:
missing value where TRUE/FALSE needed`

traceback() gives the following:

10: stop(remote_install_error(remotes[[i]], e))
9: value[3L]
8: tryCatchOne(expr, names, parentenv, handlers[[1L]])
7: tryCatchList(expr, classes, parentenv, handlers)
6: tryCatch(res[[i]] <- install_remote(remotes[[i]], ...), error = function(e) {
stop(remote_install_error(remotes[[i]], e))
})
5: install_remotes(remotes, auth_token = auth_token, host = host,
dependencies = dependencies, upgrade = upgrade, force = force,
quiet = quiet, build = build, build_opts = build_opts, build_manual = build_manual,
build_vignettes = build_vignettes, repos = repos, type = type,
...)
4: force(code)
3: withr::with_path(rtools_path(), code)
2: pkgbuild::with_build_tools({
ellipsis::check_dots_used(action = getOption("devtools.ellipsis_action",
rlang::warn))
{
remotes <- lapply(repo, github_remote, ref = ref, subdir = subdir,
auth_token = auth_token, host = host)
install_remotes(remotes, auth_token = auth_token, host = host,
dependencies = dependencies, upgrade = upgrade, force = force,
quiet = quiet, build = build, build_opts = build_opts,
build_manual = build_manual, build_vignettes = build_vignettes,
repos = repos, type = type, ...)
}
}, required = FALSE)
1: devtools::install_github("saeyslab/nichenetr")

I am using R version 3.6.3 on windows with 32gb RAM, and I have installed + used nichenet before so I am sure it's not nichenet's problem, more like something with devtools. Is there a .tar source code file where I can install nichenet directly? Thanks again.

from nichenetr.

I think you can just clone this github repository to have a local version of the package directory.

Did you also try installing it while running Rstudio or the R console as administrator?
And can you check whether you can install another package from github?

from nichenetr.

Installing other packages with devtools also gave the same error, so I'm fairly sure it is devtools problem by now.

Installing by cloning the git repository gave me a folder called 'nichenetr-master' in my library folder with all the files in the git repos, but the package is not actually installed...ie. library(nichenetr) is not recognized....

from nichenetr.

Hi @aaronkwc,

In this nichenetr-master directory (after unzipping), you can open the Rproject file nichenetr.Rproj via R studio and there click on Build --> Clean and Rebuild. Then the package should get installed.
Then close Rstudio and open the analyses scripts you have and see whether you can now load the package.

from nichenetr.

Hi @browaeysrobin ,

Worked like a charm. Somehow reinstalling nichenetr removed the problem of having inconsistent results. Now the output of my Windows machine is consistent with the Macbook, which is nice. Although the root of the problem is still unclear, I can't reproduce it now after the update so I guess I will have to leave it be. Thanks for your patience and all the help, really appreciated it.

from nichenetr.

Great to hear it worked now like it should!

from nichenetr.