ALS

Quality check

par(las=2, font=5, bty='l', ps=6)
boxplot(dat[,-2], cex=0.5, pch=16, col=runif(nrow(dat), 1, nrow(dat)))

##basic accessment of the data
###fivenum, mean, sd, missing, isNumeric
cutoff=7 #outlier pickup

SUMmat=matrix(0, ncol(dat), 10)
colnames(SUMmat)=c("minimum", "lower-hinge", "median", "upper-hinge", "maximum", "mean", "sd", "missingRate", "isNumeric", "CaseOnly")
rownames(SUMmat)=colnames(dat)
for(i in 1:nrow(SUMmat)) {
  if(is.numeric(dat[,i])) {
    SUMmat[i,1:5]=fivenum(dat[,i])
    SUMmat[i,6]=mean(dat[,i], na.rm = T)
    SUMmat[i,7]=sd(dat[,i], na.rm = T)
    naIdx=which(!is.na(dat[,i]))
    SUMmat[i,8]=1-length(naIdx)/nrow(dat)
    SUMmat[i,9]=T
    SUMmat[i,10]=ifelse(length(unique(dat[naIdx,1]))==1, T, F)

    idx=which(dat[,i] > SUMmat[i,6]+cutoff*SUMmat[i,7])
    if(length(idx)>0) {
      for(j in 1:length(idx)) {
        print(paste0("sample id: ",dat[idx[j],2], ", col:", colnames(dat)[i], ", ", "outlier value: ",  dat[idx[j], i]))
      }

    }
  } else {
    SUMmat[i,1:8]=NA
    SUMmat[i,9]=F
  }
}

## [1] "sample id: 1591411900, col:VO2HRVMAX, outlier value: 84"
## [1] "sample id: 1295458900, col:VO2KGATVMAX, outlier value: 1344.83"

write.table(SUMmat, "Als_sumDat.txt", row.names = T, col.names = T, quote = F)



basicSum=read.table("Als_sumDat.txt", as.is = T, header = T)
knitr::kable(basicSum, caption = "Data quality") %>%
kable_styling("striped", full_width = T)

Data quality
	minimum	lower.hinge	median	upper.hinge	maximum	mean	sd	missingRate	isNumeric	CaseOnly
casectrl	0.0000000	0.0000000	0.0000000	1.0000000	1.0000000	0.3827160	0.4870531	0.0000000	1	0
id	NA	NA	NA	NA	NA	NA	NA	NA	0	0
sex	0.0000000	0.0000000	1.0000000	1.0000000	1.0000000	0.5596708	0.4974512	0.0000000	1	0
age	20.0000000	44.0000000	53.0000000	60.0000000	81.0000000	52.1481481	11.4640413	0.0000000	1	0
height	143.0000000	160.0000000	166.0000000	172.0000000	187.0000000	166.0823045	8.0869127	0.0000000	1	0
weight	38.4000000	58.9000000	67.5000000	75.9000000	114.9000000	67.9300412	12.4048629	0.0000000	1	0
VO2REST	21.0000000	234.5000000	279.0000000	348.5000000	796.0000000	300.0905172	107.9820632	0.0452675	1	0
VCO2REST	83.0000000	206.0000000	254.5000000	316.5000000	665.0000000	269.6206897	97.2900920	0.0452675	1	0
VO2KGREST	1.5000000	3.5500000	4.2000000	5.0000000	9.3000000	4.4482759	1.3482381	0.0452675	1	0
RERREST	0.6300000	0.8200000	0.8900000	0.9500000	1.2800000	0.8856897	0.1102632	0.0452675	1	0
HRREST	12.0000000	75.0000000	85.0000000	93.0000000	128.0000000	84.6551724	13.7161096	0.0452675	1	0
SBPREST	3.0000000	114.0000000	125.0000000	138.0000000	191.0000000	126.1104294	20.4387088	0.3292181	1	0
DBPREST	40.0000000	72.0000000	79.0000000	87.0000000	119.0000000	78.7818182	12.1890959	0.3209877	1	0
WAT	12.5000000	50.0000000	68.0000000	91.0000000	181.0000000	73.0070423	29.6154791	0.1234568	1	0
VO2KGAT	1.9000000	10.7000000	13.4500000	15.1000000	36.1000000	13.5531651	4.4329443	0.1028807	1	0
RERAT	0.7800000	0.9700000	1.0000000	1.0200000	1.2200000	0.9964220	0.0549910	0.1028807	1	0
HRAT	1.2200000	104.0000000	115.0000000	127.0000000	216.0000000	116.3909174	21.9256320	0.1028807	1	0
SBPAT	99.0000000	127.0000000	139.0000000	155.5000000	215.0000000	142.6838235	23.2782532	0.4403292	1	0
DBPAT	43.0000000	71.0000000	79.0000000	86.0000000	123.0000000	78.5985294	12.9230313	0.4403292	1	0
WVMAX	0.0000000	74.5000000	108.0000000	136.0000000	275.0000000	105.6578723	54.8318242	0.0329218	1	0
VO2KGVMAX	1.1600000	16.5000000	20.2000000	23.4500000	56.4000000	19.8488066	8.1841514	0.0000000	1	0
RERVMAX	0.6900000	1.0950000	1.1700000	1.2750000	1.5800000	1.1706173	0.1647338	0.0000000	1	0
HRVMAX	31.0000000	125.0000000	144.0000000	158.0000000	193.0000000	139.6942149	27.9383860	0.0041152	1	0
VO2HRVMAX	2.0000000	7.0000000	9.0000000	11.0000000	84.0000000	9.7568465	6.1809220	0.0082305	1	0
SBPVMAX	101.0000000	142.0000000	165.0000000	191.0000000	247.0000000	166.9902913	30.6839102	0.1522634	1	0
DBPVMAX	37.0000000	73.0000000	82.0000000	91.0000000	123.0000000	81.7004831	14.0918330	0.1481481	1	0
WVMAXP	0.0000000	61.0000000	78.0000000	95.0000000	181.0000000	74.8680851	33.9829589	0.0329218	1	0
VEVMAXP	4.0000000	31.5000000	40.0000000	48.0000000	101.0000000	39.1275720	15.4289956	0.0000000	1	0
VO2KGVMAXP	8.0000000	54.5000000	69.0000000	82.5000000	142.0000000	67.3374486	25.4879918	0.0000000	1	0
HRVMAXP	4.0000000	77.0000000	86.0000000	94.0000000	162.0000000	83.2933884	16.9732728	0.0041152	1	0
VO2HRVMAXP	13.0000000	65.0000000	79.0000000	94.0000000	163.0000000	78.4504132	25.9189532	0.0041152	1	0
VEVCO2VMAXP	52.0000000	72.0000000	85.0000000	97.0000000	190.0000000	88.2098765	21.7002777	0.0000000	1	0
WATP	12.0000000	40.0000000	48.0000000	61.0000000	102.0000000	51.3283582	15.8751895	0.1728395	1	0
VEATP	7.0000000	17.0000000	20.0000000	23.0000000	47.0000000	20.7427184	5.8046928	0.1522634	1	0
VO2KGATP	14.0000000	37.0000000	44.0000000	53.0000000	89.0000000	45.9174757	12.8938626	0.1522634	1	0
HRATP	43.0000000	62.0000000	68.0000000	75.0000000	120.0000000	69.2585366	9.8262597	0.1563786	1	0
VO2HRATP	21.0000000	56.0000000	65.0000000	76.0000000	110.0000000	66.0195122	15.9038043	0.1563786	1	0
VEVCO2ATP	49.0000000	66.0000000	76.0000000	89.0000000	142.0000000	78.7766990	16.5505476	0.1522634	1	0
VO2KGATVMAX	8.8000000	55.6500000	63.7750000	71.7300000	1344.8300000	69.7491284	87.6162654	0.1028807	1	0
VEVCO2slope	-14.9040000	24.5810000	27.3520000	30.4520000	44.3660000	27.3968118	5.6257706	0.0246914	1	0
WeberGrading	1.0000000	1.0000000	1.0000000	2.0000000	4.0000000	1.7759336	1.0122179	0.0082305	1	0
OUES	199.0190000	1184.3390000	1534.5970000	1854.1230000	4380.8380000	1556.9218692	579.3304075	0.0246914	1	0
HRRecovery	3.0000000	18.0000000	24.0000000	31.0000000	59.0000000	24.9065421	9.7451398	0.1193416	1	0
FVC	40.0000000	78.7000000	90.0000000	98.0000000	125.0000000	87.9868421	14.4281053	0.0617284	1	0
FEV1	33.0000000	79.5000000	91.0000000	99.0000000	121.0000000	88.9464912	14.8768553	0.0617284	1	0
FEV1FVC	43.0000000	76.0000000	81.1300000	88.0000000	113.0000000	83.0824123	11.1375318	0.0617284	1	0
FEF25	15.0000000	66.0000000	86.0000000	107.0000000	162.0000000	86.5184211	28.1980459	0.0617284	1	0
FEF75	12.0000000	59.5000000	75.5000000	99.5000000	268.0000000	81.1789474	33.1258353	0.0617284	1	0
FEF2575	13.0000000	71.0000000	93.0000000	110.0000000	233.0000000	92.9543860	29.4742915	0.0617284	1	0
FEFMAX	15.0000000	59.0000000	80.0000000	97.0000000	135.0000000	78.1254386	23.7280214	0.0617284	1	0
FIVC	1.2200000	2.2350000	2.8300000	3.7850000	6.4300000	3.0547807	1.0066276	0.0617284	1	0
FIFMAX	0.7500000	1.8850000	2.6300000	3.7850000	8.5700000	2.8509649	1.2723709	0.0617284	1	0
MVV	12.0000000	73.0000000	88.0000000	101.0000000	145.0000000	86.1951220	23.5890257	0.1563786	1	0
SVC	11.0000000	78.0000000	88.0000000	97.0000000	124.0000000	86.9103774	15.8815666	0.1275720	1	0
IC	22.0000000	76.0000000	88.0000000	102.0000000	150.1000000	88.3240175	20.4364374	0.0576132	1	0
ERV	16.0000000	60.5000000	85.0000000	113.5000000	283.0000000	89.3622807	40.6693466	0.0617284	1	0
IntracellularWater	13.3000000	18.2000000	21.0000000	25.5000000	34.7000000	21.7287554	4.4329766	0.0411523	1	0
ExtracellularWater	8.5000000	11.4000000	13.2000000	15.5000000	21.5000000	13.5682403	2.5948764	0.0411523	1	0
Protein	5.8000000	7.9000000	9.1000000	11.0000000	15.0000000	9.3957082	1.9181480	0.0411523	1	0
InorganicSalt	2.1800000	2.8700000	3.2600000	3.7900000	5.3200000	3.3236052	0.6241835	0.0411523	1	0
FatMass	3.7000000	16.0000000	19.6000000	24.0000000	44.3000000	20.0991416	6.2245620	0.0411523	1	0
SoftLeanMass	28.2000000	38.0000000	43.6000000	52.7000000	72.1000000	45.2669528	9.0400249	0.0411523	1	0
FatFreeMass	30.1000000	40.4000000	46.3000000	55.7000000	76.5000000	48.0141631	9.5347692	0.0411523	1	0
BodyWater	22.0000000	29.6000000	34.0000000	40.9000000	56.2000000	35.2969957	7.0095772	0.0411523	1	0
SkeletalMuscleMass	15.4000000	21.7000000	25.3000000	31.2000000	43.3000000	26.3386266	5.7828791	0.0411523	1	0
SMM	66.0000000	90.5000000	97.6000000	104.8000000	130.3000000	97.3751073	10.4709359	0.0411523	1	0
FM	18.6000000	155.6000000	199.9000000	248.3000000	423.7000000	202.9429185	69.4770525	0.0411523	1	0
BMI	14.3000000	22.5000000	24.4000000	26.6000000	36.3000000	24.5935622	3.3806883	0.0411523	1	0
PercentBodyFat	8.7000000	24.8000000	29.8000000	34.4000000	50.7000000	29.3763948	6.9291489	0.0411523	1	0
WaistHipRatio	0.7100000	0.8500000	0.8800000	0.9200000	0.9700000	0.8819742	0.0472289	0.0411523	1	0
VisceralFatArea	11.7000000	68.1000000	91.8000000	115.9000000	211.1000000	95.1253219	36.3052671	0.0411523	1	0
RA	1.1200000	1.9800000	2.4800000	3.1000000	4.4400000	2.5562661	0.7194540	0.0411523	1	0
RAp	51.4000000	87.3000000	101.0000000	111.9000000	161.0000000	100.5742489	17.9186049	0.0411523	1	0
RAIp	56.1000000	88.1000000	97.3000000	106.6000000	147.2000000	97.0832618	13.7671092	0.0411523	1	0
LA	1.0100000	1.9400000	2.4600000	3.1000000	4.3400000	2.5177253	0.7070296	0.0411523	1	0
Lap	52.1000000	87.8000000	98.7000000	111.6000000	150.5000000	99.1072961	17.5153944	0.0411523	1	0
LAIp	56.2000000	87.3000000	96.2000000	104.4000000	136.3000000	95.6055794	13.3031955	0.0411523	1	0
TK	12.4700000	17.9300000	21.1500000	24.7800000	32.3600000	21.4240343	4.3165353	0.0411523	1	0
TKp	67.0000000	92.0000000	100.5000000	107.0000000	124.1000000	99.9231760	10.9719499	0.0411523	1	0
TKIp	73.6000000	92.2000000	97.8000000	102.1000000	112.3000000	96.9824034	7.3922717	0.0411523	1	0
RL	4.1600000	6.1100000	7.0900000	8.5700000	11.8900000	7.3843777	1.6271335	0.0411523	1	0
RLp	72.1000000	93.0000000	98.3000000	104.9000000	128.9000000	98.3862661	9.5682116	0.0411523	1	0
RLIp	72.6000000	91.6000000	95.9000000	100.0000000	114.7000000	95.4416309	7.1856557	0.0411523	1	0
LL	4.0900000	6.1100000	7.1000000	8.5700000	11.7800000	7.3599142	1.6045141	0.0411523	1	0
LLp	71.9000000	93.0000000	97.7000000	104.0000000	127.8000000	98.0433476	9.0989797	0.0411523	1	0
LLIp	71.3000000	91.7000000	95.6000000	99.3000000	117.3000000	95.1429185	7.0480955	0.0411523	1	0
RAF	0.3190000	0.3310000	0.3340000	0.3380000	0.3530000	0.3347639	0.0056530	0.0411523	1	0
LAF	0.3230000	0.3310000	0.3350000	0.3380000	0.3540000	0.3351545	0.0052245	0.0411523	1	0
TKF	0.3210000	0.3330000	0.3370000	0.3420000	0.3590000	0.3374464	0.0071732	0.0411523	1	0
RLF	0.3190000	0.3340000	0.3380000	0.3440000	0.3710000	0.3391588	0.0091579	0.0411523	1	0
RAW	0.3650000	0.3780000	0.3810000	0.3850000	0.4010000	0.3816481	0.0059252	0.0411523	1	0
LAW	0.3690000	0.3780000	0.3820000	0.3850000	0.4020000	0.3820815	0.0055279	0.0411523	1	0
TKW	0.3670000	0.3800000	0.3840000	0.3900000	0.4070000	0.3844678	0.0075595	0.0411523	1	0
RLW	0.3650000	0.3800000	0.3850000	0.3920000	0.4200000	0.3862876	0.0096158	0.0411523	1	0
ECFTBF	0.3220000	0.3340000	0.3370000	0.3430000	0.3590000	0.3379442	0.0070075	0.0411523	1	0
ECWTBW	0.3680000	0.3800000	0.3840000	0.3900000	0.4070000	0.3850000	0.0073725	0.0411523	1	0
mREE	457.9164878	1633.1976470	1957.6915385	2418.6773550	5522.9295480	2112.3070184	753.1221338	0.0452675	1	0
cREE	1033.6780000	1278.9590000	1414.5950000	1590.2800000	2226.0700000	1452.4547407	232.0336696	0.0000000	1	0
REEp	0.3300000	1.1650000	1.3950000	1.6400000	3.0800000	1.4452155	0.4214980	0.0452675	1	0
AgeOfOnset	24.0000000	43.0000000	53.0000000	61.0000000	75.0000000	52.0430108	11.2190773	0.6172840	1	1
SiteOfOnset	1.0000000	2.0000000	2.0000000	2.0000000	4.0000000	2.1290323	0.8995869	0.6172840	1	1
DelayDays	60.0000000	225.0000000	334.0000000	498.0000000	1848.0000000	417.9784946	306.7132912	0.6172840	1	1
DelayMonths	2.0000000	7.5000000	11.1333333	16.6000000	61.6000000	13.9322581	10.2232808	0.6172840	1	1
heightpre	143.0000000	160.0000000	167.0000000	172.0000000	183.0000000	165.8172043	7.6142690	0.6172840	1	1
weightpre	49.0000000	56.5000000	67.0000000	75.0000000	115.0000000	67.7010870	12.2850887	0.6213992	1	1
weightdiagnosis	37.0000000	55.5000000	66.0000000	72.0000000	110.0000000	65.2554348	11.8767699	0.6213992	1	1
BMIpre	0.0000000	22.7583068	23.8751148	26.4721074	36.2959222	24.2881376	4.2378768	0.6172840	1	1
BMIdiagnosis	0.0000000	21.6441207	23.3843537	25.8585722	34.7178387	23.4042402	4.0733476	0.6172840	1	1
ALSFRSRDiagnosis	28.0000000	39.0000000	43.0000000	45.0000000	48.0000000	42.1931818	4.1657000	0.6378601	1	1
chWeightPre	-2.8571429	-0.3861435	0.0000000	0.0000000	0.4615385	-0.2662982	0.5416928	0.6213992	1	1
chBMIPre	-0.1666173	0.0000000	0.0000000	0.1529955	0.9657730	0.0958385	0.1898371	0.6213992	1	1
chFSPre	0.0000000	0.1750628	0.4216035	0.7183535	2.4161074	0.5358948	0.5050161	0.6378601	1	1

#row_spec(row=16:16, color="white", background="red")

Feature selection

Smat=matrix(0, ncol(dat)-2, 4)

for(i in 3:ncol(dat)) {
  mod=glm(dat[,1]~dat[,i], family = "binomial")
  Smat[i-2,]=summary(mod)$coefficients[2,]
}
rownames(Smat)=colnames(dat)[3:ncol(dat)]


aucMat=matrix(2, ncol(dat)-2, 1)
rownames(aucMat)=colnames(dat)[3:ncol(dat)]
for(i in 3:ncol(dat)) {
  if(!SUMmat[colnames(dat)[i],10]) {
    rocObj=roc(dat[,1], dat[,i], quiet = T)
    aucMat[i-2,1]=rocObj$auc
  }
}

layout(matrix(c(1, 2, 3, 4), 2, 2, byrow = F))
par(las=2, cex=0.5, mai=c(1,0.5,0.5,0.5))
pcut=-log10(0.05/(ncol(dat)-2))
barplot(main="case-control",-log10(Smat[,4]), col=ifelse(-log10(Smat[,4])>pcut, "cyan", "grey"))
abline(h=-log10(0.05/ncol(dat)), col="red", lty=2)

barplot(main="case-control", aucMat, beside = T, col=ifelse(-log10(Smat[,4])>pcut, "cyan", "grey"))
plot(-log10(Smat[,4]), pch=16,aucMat,col=ifelse(-log10(Smat[,4])>pcut, "cyan", "grey"))
abline(v=pcut, col="red", lty=2)

###########
SVCmat=matrix(1, ncol(dat)-2, 4)
for(i in 3:ncol(dat)) {
  if(i!=68) {
    mod=lm(dat[,68]~dat[,i])
    if(nrow(summary(mod)$coefficients)>1) {
      SVCmat[i-2,]=summary(mod)$coefficients[2,]
    }
  }
}
rownames(SVCmat)=colnames(dat)[3:ncol(dat)]

par(las=2, cex=0.5, mai=c(1.5,0.5,0.5,0.5))
pcut=-log10(0.05/(ncol(dat)-2))
barplot(-log10(Smat[,4]), col=ifelse(-log10(Smat[,4])>pcut, "cyan", "grey"))
abline(h=-log10(0.05/ncol(dat)), col="red")

ALS

Chen Guo-Bo [chenguobo@gmail.com]

2020-07-24

Table of contents

Quality check

Feature selection