OSCR

Plasma proteomics reveals molecular overlap between physical activity and dementia risk.

Code ↔ Paper

The paper beside its authors' code: matches between them have not been computed for this paper yet.

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

R · 295 lines · 17 KB · MIT

  1. MAGMA.SPA <- function(dummyVar="",env=.GlobalEnv) {
  3. suppressPackageStartupMessages(require(WGCNA,quietly=TRUE))
  4. suppressPackageStartupMessages(require(statmod,quietly=TRUE))
  5. suppressPackageStartupMessages(require(xlsx,quietly=TRUE))
  6. suppressPackageStartupMessages(require(ggplot2,quietly=TRUE))
  7. suppressPackageStartupMessages(require(gridBase,quietly=TRUE))
  8. suppressPackageStartupMessages(require(grid,quietly=TRUE))
  9. suppressPackageStartupMessages(require(gplots,quietly=TRUE))
  10. suppressPackageStartupMessages(require(calibrate,quietly=TRUE))
  12. if (!exists("outFilePrefix")) { outFilePrefix="" } else { if (nchar(outFilePrefix)>0) outFilePrefix=paste0(outFilePrefix,".") }
  13. if (!exists("outFileSuffix")) { if (exists("FileBaseName")) { outFileSuffix=paste0("-",FileBaseName) } else { outFileSuffix="-unspecified_study" }} else { if (nchar(outFileSuffix)>0) outFileSuffix=paste0("-",outFileSuffix) }
  15. if (!exists("maxP")) { cat("- maxP variable for maximum p of nominally significant genes to keep for permutation not set. Defaulting to maxP=0.05 ...\n"); maxP=0.05; }
  16. if (!exists("FDR")) { cat("- FDR variable for q value of significant enrichment of risk in any module not set. Defaulting to FDR=0.10 ...\n"); FDR=0.10; }
  17. if (!exists("barcolors")) { cat("- barcolors variable not set. Default colors will be used...\n"); barcolors=c("darkslateblue","hotpink","mediumorchid","seagreen3","skyblue","goldenrod","darkorange","darkmagenta","darkred","darkgreen","darkturquoise","saddlebrown","maroon","honeydew","coral","purple","orangered3","lightcoral","cyan","yellow")[1:length(MAGMAinputs)]; }
  19. if (!exists("relatednessOrderBar") | !is.logical("relatednessOrderBar")) { cat("- relatednessOrderBar not found or not TRUE/FALSE. Ordering modules by relatedness order in MEs.\n"); relatednessOrderBar=TRUE; }
  20. if (!exists("plotOnly")) { cat("- plotOnly not found. Performing all calculations on provided inputs anew.\n"); plotOnly=FALSE; }
  21. if (!is.logical(plotOnly)) { cat("- plotOnly not TRUE/FALSE. Performing all calculations on provided inputs anew.\n"); plotOnly=FALSE; }
  23. if (!plotOnly) {
  24. if (!exists("cleanDat")) stop("\ncleanDat variable must exist, holding gene product (rows) X sample (columns) data in the form of log2(relative abundance).\n\n")
  26. if (!exists("NETcolors")) if(exists("net")) { if ("colors" %in% names(net)) { NETcolors=net$colors } else { NETcolors=c() } } else { NETcolors=c() }
  27. if (!length(NETcolors)==nrow(cleanDat)) { stop("\nNetwork color assignment vector not supplied or not of length in rows of cleanDat; will not be included in output table and data frame.\n\n") }
  29. MAGMAinputDir=gsub("\\/\\/","/",paste0(MAGMAinputDir,"/"))
  30. for (file in MAGMAinputs) if (!file.exists(paste0(MAGMAinputDir,file))) stop(paste0("\n",MAGMAinputDir,file," not found. Cannot continue.\n\n"))
  32. if("MEs" %in% names(net)) MEs=net$MEs
  33. if(!exists("MEs") | !exists("calculateMEs")) calculateMEs=TRUE
  34. if(calculateMEs) {
  35. cat("- MEs data frame for module eigengenes not found or calculateMEs=TRUE.\n Attempting to recalculate from cleanDat and net$colors or NETcolors...\n")
  36. if(!exists("NETcolors")) if("colors" %in% names(net)) NETcolors=net$colors
  37. # if(!exists("cleanDat") | !exists("NETcolors")) stop("Cannot find NETcolors and/or cleanDat for ME calculation.\n")
  38. MEs<-tmpMEs<-data.frame()
  39. MEList = moduleEigengenes(t(cleanDat), colors = NETcolors)
  40. MEs = orderMEs(MEList$eigengenes)
  41. colnames(MEs)<-gsub("^ME","",colnames(MEs)) # let's be consistent in case prefix was added, remove it.
  42. if("grey" %in% colnames(MEs)) MEs[,"grey"] <- NULL
  43. }
  44. colnames(MEs)=gsub("^ME","",colnames(MEs))
  45. if("grey" %in% colnames(MEs)) MEs[,"grey"] <- NULL
  47. if(!exists("parallelThreads")) { cat("- parallelThreads variable not set to a number. Attempting to use ",as.numeric(length(MAGMAinputs))," threads for quickest processing...\n"); parallelThreads=as.numeric(length(MAGMAinputs)); }
  48. if(!is.numeric(as.numeric(parallelThreads))) { cat("- parallelThreads variable not set to a number. Attempting to use ",as.numeric(length(MAGMAinputs))," threads for quickest processing...\n"); parallelThreads=as.numeric(length(MAGMAinputs)); }
  50. cat(paste0("\nSetting up parallel backend with ",parallelThreads," threads...\n"))
  51. suppressPackageStartupMessages(require(doParallel, quietly=TRUE))
  52. # if(exists("clusterLocal")) stopCluster(clusterLocal)
  53. clusterLocal <- makeCluster(c(rep("localhost",parallelThreads)),type="SOCK")
  54. registerDoParallel(clusterLocal)
  56. moduleList=sapply( colnames(MEs),function(x) as.vector(data.frame(do.call("rbind",strsplit( paste0(data.frame(do.call("rbind",strsplit(rownames(cleanDat),"[|]")))[,1],";") ,"[;]")))[,1] )[which(unlist(NETcolors)==x)] )
  57. nModules=length(names(moduleList))
  58. for (b in 1:nModules) {
  59. moduleList[[b]] <- unique(moduleList[[b]][moduleList[[b]] != ""])
  60. moduleList[[b]] <- unique(moduleList[[b]][moduleList[[b]] != "0"])
  61. }
  63. # Order modules by size of modules [as determined using standard ranked colors (no ties)]
  64. geneList <- list()
  65. modcolors=unique(unlist(NETcolors))
  66. modcolors<-modcolors[which(!modcolors=="grey")]
  67. nModules=length(modcolors)
  68. modcolors=labels2colors(c(1:nModules)) # INSURE CORRECT RANK ORDER
  69. for (i in 1:length(modcolors)) geneList[[ modcolors[i] ]] <- moduleList[[ modcolors[i] ]]
  71. orderedLabels<- cbind(paste("M",seq(1:nModules),sep=""),labels2colors(c(1:nModules)))
  72. xlabels.rankOrder <- orderedLabels[,1]
  74. cat(paste0("Performing boostrap statistics to find mean scaled enrichment scores of significant gene-level risk in ",nModules," modules with ",length(MAGMAinputs)," lists.\n[1 list per each of up to ",parallelThreads," threads at a time]...\n"))
  75. parallel::clusterExport(cl=clusterLocal, list("maxP","MAGMAinputDir","geneList","xlabels.rankOrder","MAGMAinputs","FDR"), envir=environment()) ## avoid variable not found error during foreach below:
  76. statOutList <- foreach(thisMAGMAinputFile=as.character(MAGMAinputs)) %dopar% {
  78. ## Prepare SNP_data
  79. SNP_data <- read.csv(paste(MAGMAinputDir,thisMAGMAinputFile,sep=""),stringsAsFactors=FALSE,header=T) #read.table(genePValues, stringsAsFactors = FALSE,header=T)
  80. SNP_data <- as.data.frame(SNP_data)
  82. #convert csv input's column 2 from p value if not already -log10(p) and filter p<=maxP only
  83. if(max(SNP_data[,2],na.rm=TRUE)<=1) SNP_data[,2]= -log(SNP_data[,2])
  84. SNP_data<-SNP_data[which(SNP_data[,2]>= -log10(maxP)),]
  86. nperm <- 10000
  88. # Create function for carrying out permutations for Null distribution
  89. permute <- function(pVals_SNP,ind){
  90. pVals_permuted <- sample(pVals_SNP)
  91. mean(pVals_permuted[ind])
  92. }
  94. ## Permutation p Values can be zero when the significance of the association is very high. The following package (statmod) has a function permp that overcomes this problem and calculates a permutation p value using method described here http://www.statsci.org/webguide/smyth/pubs/permp.pdf
  96. require(statmod)
  97. commonGenes<-list()
  98. for (i in 1:nModules){
  99. pVals_SNP <- as.numeric(SNP_data[c(1:nrow(SNP_data)),2])
  100. ind <- which(SNP_data[,1] %in% geneList[[i]])
  101. commonGenes[[i]] <- as.vector(na.omit(intersect(geneList[[i]],SNP_data[,1])))
  103. module_mean <- mean(pVals_SNP[ind])
  104. module_sd <- sd(pVals_SNP[ind])
  105. module_sem <- module_sd / sqrt(length(geneList[[i]]))
  106. permMean <- replicate(nperm,permute(pVals_SNP,ind))
  107. pVal_module <- sum(abs(permMean) >= abs(module_mean)) / nperm
  108. numCommon <- length(commonGenes[[i]])
  109. NES_perm <- (permMean - mean(permMean)) / sd(permMean)
  110. NES_module <- (module_mean - mean(permMean)) / sd(permMean)
  111. NESpVal_module <- sum(abs(NES_perm) >= abs(NES_module)) / nperm
  112. if (!is.na(NESpVal_module)){
  113. statmodP <- permp(sum(abs(NES_perm) >= abs(NES_module)),nperm,length(SNP_data[[1]]),length(geneList[[i]]))
  114. }else{
  115. statmodP <- "NA"
  116. }
  117. if (i == 1){
  118. mean_allModules <- module_mean
  119. # commonGenes_all <- c(commonGenes[[i]]) #,rep(NA,len=50)
  120. permMean_all <- permMean
  121. pVal_all <- pVal_module
  122. numCommon_all <- numCommon
  123. NES_perm_all <- NES_perm
  124. NES_module_all <- NES_module
  125. NESpVal_all <- NESpVal_module
  126. statmodP_all <- statmodP
  127. }else{
  128. mean_allModules <- c(mean_allModules,module_mean)
  129. # commonGenes_all <- cbind(commonGenes_all,c(commonGenes[[i]])) #,rep(NA,len=50)
  130. permMean_all <- cbind(permMean_all,permMean)
  131. pVal_all <- cbind(pVal_all,pVal_module)
  132. numCommon_all <- c(numCommon_all,numCommon)
  133. NES_perm_all <- cbind(NES_perm_all, NES_perm)
  134. NES_module_all <- c(NES_module_all, NES_module)
  135. NESpVal_all <- cbind(NESpVal_all,NESpVal_module)
  136. statmodP_all <- cbind(statmodP_all,statmodP)
  138. }
  140. }
  142. maxHitListSize<-max(unlist(lapply(commonGenes,length)))
  143. commonGenes_all1<-lapply(commonGenes,function(x) if (length(x)==maxHitListSize) { sort(x) } else { c(sort(x), rep(NA,maxHitListSize-length(x))) })
  144. #commonGenes_all<- as.data.frame(matrix(NA,nrow=maxHitListSize,ncol=0))
  145. #for (i in 1:length(commonGenes_all1)) commonGenes_all<-cbind(commonGenes_all, commonGenes_all1[[i]])
  146. commonGenes_all <- matrix(unlist(commonGenes_all1),nrow=maxHitListSize,ncol=length(commonGenes_all1),byrow=FALSE)
  148. names(mean_allModules) <- names(geneList)
  149. names(numCommon_all) <- names(geneList)
  150. names(NES_module_all) <- names(geneList)
  151. colnames(commonGenes_all) <- names(geneList)
  152. colnames(permMean_all) <- names(geneList)
  153. colnames(pVal_all) <- names(geneList)
  154. colnames(NES_perm_all) <- names(geneList)
  155. colnames(NESpVal_all) <- names(geneList)
  156. colnames(statmodP_all) <- names(geneList)
  158. ## Write output to tables and PDF
  159. pVal_names<-colnames(pVal_all)
  160. pVal_all<-as.vector(pVal_all)
  161. names(pVal_all)<-pVal_names
  163. NESpVal_names<-colnames(NESpVal_all)
  164. NESpVal_all<-as.vector(NESpVal_all)
  165. names(NESpVal_all)<-NESpVal_names
  167. statmodP_names<-colnames(statmodP_all)
  168. statmodP_all<-as.vector(statmodP_all)
  169. names(statmodP_all)<-statmodP_names
  171. all_output <- rbind(mean_allModules, NES_module_all, numCommon_all,
  172. pVal_all, NESpVal_all, statmodP_all,
  173. commonGenes_all)
  174. # write.table(all_output,file=paste("./",outFilePrefix,"MAGMA-SPA-",thisMAGMAinputFile,".txt",sep=""), na="", row.names=T,col.names=NA,sep="\t", quote=FALSE)
  176. pdf(paste("./",outFilePrefix,"MAGMA-SPA-",thisMAGMAinputFile,".pdf",sep=""),height=8,width=8)
  177. par(mfrow=c(4,4))
  178. par(mar=c(4,4,4,2))
  179. barplot(mean_allModules,main = "Mean - Enrichment Score", ylab = "Mean",cex.names=0.8, width=0.8,las=2,cex.main=0.95,names.arg=xlabels.rankOrder)
  180. barplot(NES_module_all,main = "Mean - Scaled Enrichment Score", ylab = "Mean",cex.names=0.8, width=0.8,las=2,cex.main=0.95,names.arg=xlabels.rankOrder)
  182. modColors <- names(geneList)
  183. for (a in 1:nModules){
  184. modClr <- xlabels.rankOrder[a]
  185. NESpValue_module <- NESpVal_all[a]
  186. if (is.na(NESpValue_module)){
  187. next
  188. } else{
  189. par(mar=c(4,4,4,4))
  190. permMean_Module <- NES_perm_all[,a]
  191. hist(permMean_Module,xlab = "Normalized ES (Mean)", ylab = "Frequency", main=modClr, sub = paste("pValue",NESpValue_module),col="grey",freq=T, border="black",cex.main=0.95)
  192. abline(v=NES_module_all[a],col="red")
  193. }
  194. }
  195. dev.off()
  198. return(list(all_output, NES_module_all,thisMAGMAinputFile))
  199. }
  201. # re-combine list elements from two outputs, from all MAGMAinputFile runs
  202. all_output = do.call(list, lapply(statOutList,function(x){x[[1]]}) )
  203. NES_module_all = do.call(list, lapply(statOutList,function(x){x[[2]]}) )
  204. names(all_output)<-names(NES_module_all) <- MAGMAinputs <- do.call(c,lapply(statOutList,function(x){x[[3]]}))
  206. all_output <- lapply(all_output,function(x) { rownames(x)[7:nrow(x)]<- paste0("geneHit",1:(nrow(x)-6)); x; })
  208. ##Export all gene lists to multi-sheet Excel
  209. my.xlsx=createWorkbook()
  210. for (sheetName in MAGMAinputs) addDataFrame(as.data.frame(all_output[[sheetName]]), sheet=createSheet(my.xlsx, sheetName), startColumn=1, row.names=TRUE,col.names=TRUE)
  211. saveWorkbook(my.xlsx, paste0("./",outFilePrefix,"MAGMA-SPA",outFileSuffix,".xlsx"))
  214. #################
  215. ## Summary plot of multiple GWAS MAGMA bootstrap statistics
  217. allBarData<-do.call(rbind,NES_module_all) #(AD.IGAP.1234=AD1234_NES_all,AD.2019=AD.Kunkle_NES_all,ASD=ASD_NES_all,SCZ=SCZ_NES_all,PD=PD_NES_all,ALS=ALS_NES_all)
  218. rownames(allBarData)<-MAGMAinputs
  219. allBarData<-allBarData[,c(1:nModules)]
  220. allBarData[!is.finite(allBarData)]<-0
  222. xlabels = if (relatednessOrderBar) { orderedLabels[match(colnames(MEs),orderedLabels[,2]),1] } else { xlabels.rankOrder }
  223. # handle single GWAS list input case, and reorder module bars if relatednessOrderBar==TRUE # Feb 2, 2023 bugfix.
  224. if (as.numeric(length(MAGMAinputs))==1) { allBarData<-data.frame(oneInput=allBarData[if (relatednessOrderBar) { colnames(MEs) } else { 1:ncol(MEs) }]); colnames(allBarData)=as.character(MAGMAinputs)[1]; allBarData=t(allBarData); } else { if (relatednessOrderBar) allBarData <- allBarData[,colnames(MEs)] }
  226. } #end if (!plotOnly)
  228. # Check that plot data exists, in case plotOnly==TRUE
  229. if (!exists("xlabels") | !exists("allBarData")) stop("Processed MAGMA Enrichment in Modules not found. Cannot plot.\nRerun with plotOnly=FALSE.\n\n")
  231. ##Simplest output
  232. # pdf(paste0("./",outFilePrefix,"MAGMA-Enr_BarPlot",outFileSuffix,".pdf"),height=8,width=16)
  233. # par(mfrow=c(1,1))
  234. # par(mar=c(4,6,4,2))
  235. # barplot(allBarData,main =paste0("Enrichment of MAGMA-implicated Genetic Risk of Disease(s)\nbased on ",length(MAGMAinputs)," GWAS-derived MAGMA summary p (<=",maxP,") gene lists,\n(as published in Seyfried, et al, Cell Systems, 2017)"), ylab = "Mean-scaled Enr. Score",cex.names=0.70, cex.lab=1.75, width=0.8,las=2,cex.main=0.95,
  236. # names.arg=xlabels, beside=TRUE,
  237. # col=barcolors,
  238. # legend.text=TRUE, args.legend=list(x="top", bty="n", inset=c(0, 0)), xpd=FALSE,ylim=c(min(na.omit(allBarData))-1,max(na.omit(allBarData))+1.3))
  239. # abline(h=1.28, lty=2, col="red")
  240. # dev.off()
  242. ## Output with inset for color swatches for x axis labels
  244. ## Create ordered swatches for inset at x axis using ggplot bar
  245. colorData=data.frame(Mnum=as.character(xlabels),yBlank=c(1), fill=gplots::col2hex(unique(WGCNA::labels2colors(as.numeric(gsub("M","",as.character(xlabels)))))), fillName=unique(WGCNA::labels2colors(as.numeric(gsub("M","",as.character(xlabels))))))
  246. # barplot(as.matrix(diff(1:93)), horiz=T, col=colorData$fillName, axes=F, xlab=NA)
  247. p <- ggplot(colorData, aes(x=Mnum, y=yBlank)) + geom_bar(stat="identity", fill=colorData$fill, color="#000000", size=0.01, width = 0.8675, aes(fill=fillName)) +theme(axis.title = element_blank(), axis.text.x = element_text(face="bold", color="#000000", size=12, angle=90), axis.ticks.y=element_blank(), panel.background = element_rect(fill = "transparent",colour = NA), panel.grid.minor = element_blank(), panel.grid.major = element_blank(), plot.background = element_rect(fill = "transparent",colour = NA)) +
  248. scale_x_discrete( limits=colorData$Mnum) + theme(axis.text.x=element_text(angle=90, hjust=1, vjust=0.3), axis.text.y=element_text(color="#FFFFFF")) + labs(x="", y="") + scale_y_continuous(expand = expansion(mult = c(0, 0)), limits=c(0,1), breaks=c(0), label=c(""))
  251. pdf(file=paste0("./",outFilePrefix,"MAGMA-Enr_BarPlot",outFileSuffix,".pdf"),width=16,height=9, onefile=FALSE) # onefile=FALSE for forcing no blank page 1.
  253. grid.newpage()
  254. # grid.rect(gp=gpar(col="black"))
  255. #grid.text("",
  256. # y=unit(1, "npc") - unit(1, "lines"), gp=gpar(col="black"))
  257. #grid.rect(gp=gpar(col="green"))
  259. pushViewport(viewport(width=1, height=1, name="OuterFrame", just=c("center","center")))
  261. par(mar=c(5,6,4,2))
  262. plot.new()
  264. # par(omi = gridFIG(), new = TRUE)
  265. par(fig = gridFIG(), new = TRUE)
  266. opar <- par(lwd = 0.01)
  268. barplot(allBarData,main =paste0("Enrichment of MAGMA-implicated Genetic Risk of Disease(s)\nbased on ",length(MAGMAinputs)," GWAS-derived MAGMA summary p (<=",maxP,") gene lists,\n(as published in Seyfried, et al, Cell Systems, 2017)"), ylab = "Mean-scaled Enr. Score",cex.names=0.70, cex.lab=1.75, width=0.8,las=2,cex.main=0.95,
  269. names.arg=rep("",length(xlabels)), #xlabels,
  270. beside=TRUE, space=c(0,rep(c(rep(0,length(MAGMAinputs)-1),0.15*length(MAGMAinputs)),length(xlabels)-1),rep(0,length(MAGMAinputs)-1)), #border=NA,
  271. col=barcolors,
  272. legend.text=TRUE, args.legend=list(x="top", bty="n", inset=c(0, 0)), xpd=FALSE,ylim=c(min(na.omit(allBarData))-1,max(na.omit(allBarData))+1.3))
  273. abline(h=qnorm(FDR,lower=F), lty=3, lwd=1, col="red")
  274. calibrate::textxy(X=0,Y=qnorm(FDR,lower=F),paste0("FDR = ",FDR*100,"%"),cex=1)
  275. #upViewport()
  277. #For onscreen: pushViewport(viewport(width=0.92245, height=0.10, x=unit(9.255, "inch"), y=0.035)) #ok for out to screen,
  278. #For 93 modules:
  279. #pushViewport(viewport(width=0.8498, height=0.08, x=unit(8.381, "inch"), y=0.0925)) #for out to PDF
  280. #For 23 modules:
  281. #pushViewport(viewport(width=0.8618, height=0.08, x=unit(8.381, "inch"), y=0.0925)) #for out to PDF
  282. #Automatic adjustment of viewport inset for swatches (works for 43-93 modules)
  283. #pushViewport(viewport(width=0.879-0.00031*length(xlabels), height=0.08, x=0.524, y=0.0925, just=c("center","center"))) #for out to PDF
  284. #(works for 13, 43, 93 modules)
  285. pushViewport(viewport(width=0.879-(0.00031-0.00000571*(length(xlabels)-93))*length(xlabels), height=0.08, x=0.5235, y=0.0925, just=c("center","center"))) #for out to PDF
  286. par(fig = gridFIG(), new = TRUE)
  287. # just="bottom", name="Swatches"))
  288. # grid.rect(gp=gpar(col="blue"))
  289. print(p, newpage=FALSE)
  291. dev.off()
  293. return(list(allBarData=allBarData,xlabels=xlabels, all_output=all_output))
  294. } # end Function MAGMA.SPA

MAGMA.SPA.R at commit eae8988, under MIT · at the source

Overview

Authors: Rowan Saloner1, Emily W Paolillo1, Anna M VandeBunte1, Claire J Cadwallader1, Coty Chen1, Brian T Steffen2, David A Bennett3, Bradley F Boeve4, Howard J Rosen1, Adam L Boxer1, Joel H Kramer1, Kaitlin B Casaletto1
ORCID iDs: Rowan Saloner, Coty Chen, David A Bennett, Bradley F Boeve, Kaitlin B Casaletto
  1. Memory and Aging Center, Department of Neurology, Weill Institute for Neurosciences, University of California San Francisco, San Francisco, CA 94158, USA
  2. Division of Computational Health Sciences, Department of Surgery, University of Minnesota, Minneapolis, MN 55455, USA
  3. Rush Alzheimer’s Disease Center, Rush University Medical Center, Chicago, IL 60612, USA
  4. Department of Neurology, Mayo Clinic, Rochester, MN 55905, USA
Institutions: University of California, San Francisco (United States); University of Minnesota (United States); Rush University Medical Center (United States); Mayo Clinic (United States)
Journal: Brain communications, volume 8, issue 4, article fcag287
Dates: received 3 December 2025; accepted 25 June 2026; published online 21 July 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1093/braincomms/fcag287 · PMID 42540732 · PMCID PMC13426312 · OpenAlex W7169833032
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: genetics / omics (modality), human (organism), Alzheimer's / dementia (population)
Methods: Statistics, Preprocessing, Connectivity, fMRI & imaging
Keywords: exercise, angiogenesis, dementia, proteomics, biomarkers
Topic: Dementia and Cognitive Impairment Research (Psychiatry and Mental health, Medicine), according to OpenAlex
Funding: NIH (R01AG032289, R01AG048234, P30AG062422, R01AG072475, UF1NS100608, K23AG058752, K23AG090757, K23AG084883, P50AG047366, P30AG066515, P30AG10161, P30AG72975, R01AG15819, R01AG17917, U01AG46152, U01AG61356); Alzheimer’s Association (AARG-20-683875, AARF-23-1145318, AARF-22-974065); Larry L. Hillblom Foundation (2024-A-001-CTR, 2018-A-006-NET); New Vision Research (CCAD 2024-001-1); American Academy of Neurology; Association for Frontotemporal Lobar Degeneration; American Brain Foundation; Shenandoah Foundation; Advancing Research and Treatment for FrontoTemporal Lobar Degeneration (U54NS092089); NINDS; National Center for Advancing Translational Sciences; Longitudinal Evaluation of Familial Frontotemporal Dementia Subjects (U01AG045390); National Institute on Aging; ARTFL/LEFFTDS Longitudinal Frontotemporal Lobar Degeneration (U19AG063911); Gates Ventures; HHS | NIH | National Heart, Lung, and Blood Institute (NHLBI) (R01HL134320); Department of Health and Human Services (75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, 75N92022D00005); SomaLogic Inc
Citations: cited by 1 paper (Europe PMC); 55 references in the paper

Abstract

Physical activity (PA) is a modifiable lifestyle behaviour associated with lower dementia risk; however, molecular pathways bridging PA-related dementia prevention are poorly understood. We leveraged large-scale plasma proteomics to identify biological signatures of objectively monitored PA and cognitive ageing in functionally intact older adults, cross-validated these signatures in independent exercise cohorts and tested associations with both symptomatic and presymptomatic stages of neurodegeneration across multiple Alzheimer’s disease and related dementias (ADRD) cohorts. We analysed large-scale plasma proteomics data (SomaScan 7k) across three cohorts including naturalistic, objective PA monitoring (University of California, San Francisco Brain Aging Network for Cognitive Health cohort, n = 65), self-reported PA (Atherosclerosis Risk In Communities study, n = 10 644) and PA intervention (Health Risk Factors, Exercise Training and Genetics study, n = 654). Differential regression models examined individual protein correlates of PA, adjusting for age and sex. Weighted gene co-expression network analysis assembled proteins into unbiased modules of protein co-expression, which were annotated for gene ontology and cell-type enrichment. To test clinical relevance to ADRD, we examined PA-related protein levels across-cohorts of symptomatic Alzheimer’s disease and Parkinson’s disease (Stanford Alzheimer’s Disease Research Center), as well as frontotemporal dementia-spectrum disorders (ARTFL/LEFFTDS Longitudinal Frontotemporal Lobar Degeneration consortium). PA-related plasma proteins were also tested as predictors of antemortem cognitive change and post-mortem brain tissue mass spectrometry proteomic signatures in brain donors from the Religious Orders Study and Rush Memory and Aging Project (ROSMAP) cohort. Differential regression and network analyses identified PA plasma proteomic signatures linked to cell adhesion/extracellular matrix (ECM), immune response and lipid metabolism. Protein co-expression module M12 ECM/neurodevelopment harboured growth factor, cell adhesion and vascular remodelling proteins that (i) were positively associated with PA across exercise cohorts, (ii) positively associated with cognitive function and (iii) negatively associated with Alzheimer’s disease, Parkinson’s disease and frontotemporal dementia. Furthermore, M12 was enriched for proteins from Alzheimer’s disease risk genes and antemortem plasma abundance of anthrax toxin receptor cell adhesion molecule 2 (ANTXR2), an M12 ‘hub’ protein and top PA hit across-cohorts, forecasted longitudinal global cognitive decline and post-mortem brain tissue signatures of synaptic function and proteolysis in ROSMAP. Collectively, our integrated systems biology analysis of six independent plasma proteomic datasets facilitated discovery and validation of blood-detectable molecular signatures of PA and neurodegenerative disease, including PA-related proteins with clinical and biological relevance to early stages of disease. Circulating levels of PA-related proteins reflecting ECM biology (e.g. ANTXR2) may represent key molecular targets for dementia prevention.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repositories

Its files are read in the Code ↔ Paper reader above.

edammer/MAGMA.SPA

License: MIT
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: eae89881c51d5b1b9a7bd43e5bdeabf4bf7d6086, 4 February 2023
Languages: R (1)
Size: 4 files, 1 script
Software Heritage: not archived
Found in: the text, “Alzheimer’s disease GWAS module association”
Holds: README, license file
Not found: CITATION.cff, environment file, tests, continuous integration, documentation
Tools: ggplot2 (1 file), WGCNA (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
3 files
At the source: github.com/edammer/MAGMA.SPA

edammer

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: the link answers
Software Heritage: not checked
Found in: “Data availability”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
At the source: github.com/edammer

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 1 script, each with its path and the digest of its content;
  • no match between paragraphs and code yet;
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data availability

UCSF BrANCH data are available on request made to the UCSF Memory and Aging Center. Academic; not-for-profit investigators can request data for professional education and for research studies. Requests can be made online (https://memory.ucsf.edu/research-trials/professional/open-science). Datasets used for the analyses for the current study are also available from the corresponding author on reasonable request. Algorithms used for protein data processing and analysis are available in existing R packages and at https://github.com/edammer, as described in the Methods. ROSMAP resources can be requested at https://www.radc.rush.edu and www.synapse.org. Pre-existing data access policies for each of the ARIC parent cohort studies specify that research data requests can be submitted to each steering committee; these will be promptly reviewed for confidentiality or intellectual property restrictions and will not unreasonably be refused. Please refer to the data sharing policies of these studies. Individual-level patient or protein data may further be restricted by consent, confidentiality or privacy laws/considerations. These policies apply to both clinical and proteomic data.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 12 authors, 5 keywords, 18 funders, 54 references.

Cite

This paper

Saloner, R., Paolillo, E. W., VandeBunte, A. M., Cadwallader, C. J., Chen, C., Steffen, B. T., Bennett, D. A., Boeve, B. F., Rosen, H. J., Boxer, A. L., Kramer, J. H., & Casaletto, K. B. (2026). Plasma proteomics reveals molecular overlap between physical activity and dementia risk. Brain communications, 8(4), fcag287. https://doi.org/10.1093/braincomms/fcag287

BibTeX

@article{saloner2026plasma,
author = {Saloner, Rowan and Paolillo, Emily W and VandeBunte, Anna M and Cadwallader, Claire J and Chen, Coty and Steffen, Brian T and Bennett, David A and Boeve, Bradley F and Rosen, Howard J and Boxer, Adam L and Kramer, Joel H and Casaletto, Kaitlin B},
title = {{Plasma proteomics reveals molecular overlap between physical activity and dementia risk}},
journal = {Brain communications},
year = {2026},
month = jul,
volume = {8},
number = {4},
pages = {fcag287},
publisher = {Oxford University Press},
issn = {2632-1297},
doi = {10.1093/braincomms/fcag287},
url = {https://doi.org/10.1093/braincomms/fcag287},
pmid = {42540732},
pmcid = {PMC13426312}
}

RIS

TY - JOUR
AU - Saloner, Rowan
AU - Paolillo, Emily W
AU - VandeBunte, Anna M
AU - Cadwallader, Claire J
AU - Chen, Coty
AU - Steffen, Brian T
AU - Bennett, David A
AU - Boeve, Bradley F
AU - Rosen, Howard J
AU - Boxer, Adam L
AU - Kramer, Joel H
AU - Casaletto, Kaitlin B
TI - Plasma proteomics reveals molecular overlap between physical activity and dementia risk
T2 - Brain communications
J2 - Brain Commun
PY - 2026
DA - 2026/07/21
VL - 8
IS - 4
SP - fcag287
SN - 2632-1297
PB - Oxford University Press
DO - 10.1093/braincomms/fcag287
UR - https://doi.org/10.1093/braincomms/fcag287
LA - en
ER -

CSL-JSON

{
"id": "10.1093/braincomms/fcag287",
"type": "article-journal",
"title": "Plasma proteomics reveals molecular overlap between physical activity and dementia risk",
"container-title": "Brain communications",
"author": [
{
"family": "Saloner",
"given": "Rowan"
},
{
"family": "Paolillo",
"given": "Emily W"
},
{
"family": "VandeBunte",
"given": "Anna M"
},
{
"family": "Cadwallader",
"given": "Claire J"
},
{
"family": "Chen",
"given": "Coty"
},
{
"family": "Steffen",
"given": "Brian T"
},
{
"family": "Bennett",
"given": "David A"
},
{
"family": "Boeve",
"given": "Bradley F"
},
{
"family": "Rosen",
"given": "Howard J"
},
{
"family": "Boxer",
"given": "Adam L"
},
{
"family": "Kramer",
"given": "Joel H"
},
{
"family": "Casaletto",
"given": "Kaitlin B"
}
],
"container-title-short": "Brain Commun",
"volume": "8",
"issue": "4",
"page": "fcag287",
"DOI": "10.1093/braincomms/fcag287",
"PMID": "42540732",
"PMCID": "PMC13426312",
"ISSN": "2632-1297",
"publisher": "Oxford University Press",
"URL": "https://doi.org/10.1093/braincomms/fcag287",
"language": "en",
"issued": {
"date-parts": [
[
2026,
7,
21
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.21203/rs.3.rs-9499814/v1 [code]
Plasma proteomics link menopause timing to brain aging and dementia risk
Journal: Research Square (preprint)
In common: WGCNA, ggplot2, Alzheimer's / dementia, genetics / omics, 2 references, author Rowan Saloner
[2] doi:10.64898/2026.04.23.26351500 [code]
Plasma proteomics link menopause timing to brain aging and dementia risk
Journal: medRxiv (preprint)
In common: WGCNA, ggplot2, Alzheimer's / dementia, genetics / omics, 2 references, author Rowan Saloner
[3] doi:10.1038/s41467-026-74038-4 [code]
Semaglutide attenuates neuroinflammation in male mice.
Journal: Nature communications
In common: WGCNA, ggplot2, Alzheimer's / dementia, 5 references
[4] doi:10.1016/j.celrep.2026.117235 [code]
Integration of aged brain multi-omics reveals cross-system mechanisms underlying Alzheimer's disease heterogeneity.
Journal: Cell reports
In common: ggplot2, Alzheimer's / dementia, genetics / omics, 5 references
[5] doi:10.1038/s41380-026-03686-1 [code]
Early oligodendrocyte dysfunction signature in Alzheimer's disease: Insights from DNA methylomics and transcriptomics.
Journal: Molecular psychiatry
In common: WGCNA, ggplot2, Alzheimer's / dementia, genetics / omics, 4 references
[6] doi:10.1523/eneuro.0468-25.2026 [code]
A Multi-Network Approach Identifies Proteins Related to Dendritic Spines in Alzheimer's Disease.
Journal: eNeuro
In common: WGCNA, ggplot2, Alzheimer's / dementia, genetics / omics, 3 references
[7] doi:10.1038/s41586-026-10793-0
Cell-type signatures of Alzheimer's disease shared across population groups.
Journal: Nature
In common: Alzheimer's / dementia, genetics / omics, 5 references
[8] doi:10.1002/alz.71823 [code]
Cellular transcriptomic signatures underpinning the heterogeneity of depression in Alzheimer's disease.
Journal: Alzheimer's & dementia : the journal of the Alzheimer's Association
In common: ggplot2, Alzheimer's / dementia, genetics / omics, 4 references
[9] doi:10.1038/s41467-026-71682-8 [code]
GWAS meta-analysis of cerebrospinal fluid Alzheimer's biomarkers reveals loci regulating lipids, brain volume and autophagy.
Journal: Nature communications
In common: ggplot2, Alzheimer's / dementia, genetics / omics, 4 references
[10] doi:10.1093/braincomms/fcag326 [code]
Brain multiomic profiling identifies tau-related transcriptomic dysregulation in Alzheimer's disease.
Journal: Brain communications
In common: Alzheimer's / dementia, genetics / omics, 5 references

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.