为什么 ggplot 制作的箱线图与 Prsim 等其他软件制作的箱线图不同？
Why the boxplot made by ggplot is different to the one made by other software like Prsim?
我使用以下代码用 ggplot 制作了一个箱线图。
ggplot(data = df.08.long,
aes(x = TMT_signals, y = as.numeric(TMT_Intensities), fill = `probe.Mod.or.not(Y/N)`)) +
geom_boxplot() +
ylim(0, 2.5e3) +
theme_classic() +
theme(axis.title=element_text(size=8),
axis.text=element_text(size=10),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1))
这是我的数据框。
structure(list(Scan.number = c(10017, 10017, 10017, 10017, 10017,
10017, 10017, 10017, 10017, 13240, 13240, 13240, 13240, 13240,
13240, 13240, 13240, 13240, 27592, 27592), Sequence = c("AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK",
"AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK"), Length = c(16L, 16L,
16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L,
16L, 16L, 16L, 37L, 37L), Missed.cleavages = c(0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 1L
), Modified.sequence = c("_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_",
"_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_",
"_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_",
"_AAAYSAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK_",
"_AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK_"), probe_TMT6.Probabilities = c("",
"", "", "", "", "", "", "", "", "AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR",
"AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR",
"AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR",
"AAAY(1)SAQVQPVDGATR", "", ""), `Uniprot ID` = c("Q9H7E9", "Q9H7E9",
"Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9",
"Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9",
"Q9H7E9", "Q9H7E9", "Q00610", "Q00610"), `probe.Mod.or.not(Y/N)` = c("N",
"N", "N", "N", "N", "N", "N", "N", "N", "Y", "Y", "Y", "Y", "Y",
"Y", "Y", "Y", "Y", "N", "N"), `kinase.or.not(Y/N)` = c("N",
"N", "N", "N", "N", "N", "N", "N", "N", "N", "N", "N", "N", "N",
"N", "N", "N", "N", "N", "N"), Gene.Names = c("C8orf33", "C8orf33",
"C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33",
"C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33",
"C8orf33", "C8orf33", "C8orf33", "C8orf33", "CLTC", "CLTC"),
Charge = c(2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 4L, 4L, 4L,
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L), m.z = c(802.90499, 802.90499,
802.90499, 802.90499, 802.90499, 802.90499, 802.90499, 802.90499,
802.90499, 647.57262, 647.57262, 647.57262, 647.57262, 647.57262,
647.57262, 647.57262, 647.57262, 647.57262, 1107.5251, 1107.5251
), Score = c(86.313, 86.313, 86.313, 86.313, 86.313, 86.313,
86.313, 86.313, 86.313, 41.695, 41.695, 41.695, 41.695, 41.695,
41.695, 41.695, 41.695, 41.695, 28.532, 28.532), Retention.time = c(27.774,
27.774, 27.774, 27.774, 27.774, 27.774, 27.774, 27.774, 27.774,
35.978, 35.978, 35.978, 35.978, 35.978, 35.978, 35.978, 35.978,
35.978, 72.556, 72.556), Precursor.Intensity = c(460631.45703125,
460631.45703125, 460631.45703125, 460631.45703125, 460631.45703125,
460631.45703125, 460631.45703125, 460631.45703125, 460631.45703125,
472201.625, 472201.625, 472201.625, 472201.625, 472201.625,
472201.625, 472201.625, 472201.625, 472201.625, 388790.9296875,
388790.9296875), Localization.prob = c(NaN, NaN, NaN, NaN,
NaN, NaN, NaN, NaN, NaN, 1, 1, 1, 1, 1, 1, 1, 1, 1, NaN,
NaN), probe_TMT6.site.IDs = c("", "", "", "", "", "", "",
"", "", "308", "308", "308", "308", "308", "308", "308",
"308", "308", "", ""), TMT_signals = c("TMT126", "TMT127N",
"TMT128N", "TMT128C", "TMT129N", "TMT129C", "TMT130N", "TMT130C",
"TMT131", "TMT126", "TMT127N", "TMT128N", "TMT128C", "TMT129N",
"TMT129C", "TMT130N", "TMT130C", "TMT131", "TMT126", "TMT127N"
), TMT_Intensities = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 1824.667,
3470.869, 1691.413, 2367.219, 1895.059, 1712.427, 1529.349,
1617.825, 1677.578, 0, 0)), row.names = c(NA, -20L), class = c("tbl_df",
"tbl", "data.frame"))
这是我从 ggplot 得到的图。
这是我从 Prism 获得的箱线图。
看起来 ggplot 和 Prism 计算的中值不同。例如，ggplot 中 126 Y 的中值约为 1500，而 Prism 中 126 Y 的中值约为 5000。
我检查了我的数据集，这些列中没有 NA 值。我自己计算的中值与 Prism 中的值相匹配。
有谁知道发生了什么事吗？
谢谢！
这里的问题出在 ylim 中，它与 scale_y_continuous(limits = ...) 一起有一种让一些用户感到意外的行为。如 ?ylim、
的帮助中所述
This is a shortcut for supplying the limits argument to the individual
scales. By default, any values outside the limits specified are
replaced with NA. Be warned that this will remove data outside the
limits and this can produce unintended results. For changing x or y
axis limits without dropping data observations, see coord_cartesian().
在汇总操作期间，这会产生特别令人困惑的结果，就像您在 geom_boxplot() 中得到的那样，因为它不会出错，它只会产生不同的结果和一个您可能会错过或忽略的警告.
例如，在下图中，我们期望箱线图的范围为 0 到 100，但只得到零值。这是因为使用 ylim 或 scale_y_continuous(limits = ...) 都会在执行任何汇总计算之前过滤掉范围外的数据。
df1 <- data.frame(x = 1, y = c(0, 100))
ggplot(df1, aes(x, y)) +
geom_boxplot() +
ylim(c(0,90))
有一个警告，但很容易错过或没有意识到它的含义：
Warning message: Removed 1 rows containing non-finite values
(stat_boxplot).
这是说其中一个结果是 NA（在上述步骤之后）并且在统计计算中被忽略。
要得到预期的效果，一般应该使用coord_cartesian(ylim = ...)来控制观看window。
ggplot(df1, aes(x, y)) +
geom_boxplot() +
coord_cartesian(ylim = c(0,90))
我使用以下代码用 ggplot 制作了一个箱线图。
ggplot(data = df.08.long,
aes(x = TMT_signals, y = as.numeric(TMT_Intensities), fill = `probe.Mod.or.not(Y/N)`)) +
geom_boxplot() +
ylim(0, 2.5e3) +
theme_classic() +
theme(axis.title=element_text(size=8),
axis.text=element_text(size=10),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1))
这是我的数据框。
structure(list(Scan.number = c(10017, 10017, 10017, 10017, 10017,
10017, 10017, 10017, 10017, 13240, 13240, 13240, 13240, 13240,
13240, 13240, 13240, 13240, 27592, 27592), Sequence = c("AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR", "AAAYSAQVQPVDGATR",
"AAAYSAQVQPVDGATR", "AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK",
"AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK"), Length = c(16L, 16L,
16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L, 16L,
16L, 16L, 16L, 37L, 37L), Missed.cleavages = c(0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 1L
), Modified.sequence = c("_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_",
"_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_",
"_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_", "_AAAYSAQVQPVDGATR_",
"_AAAYSAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAAY(XO44_TMT6)SAQVQPVDGATR_",
"_AAAY(XO44_TMT6)SAQVQPVDGATR_", "_AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK_",
"_AAEQAHLWAELVFLYDKYEEYDNAIITMMNHPTDAWK_"), probe_TMT6.Probabilities = c("",
"", "", "", "", "", "", "", "", "AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR",
"AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR",
"AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR", "AAAY(1)SAQVQPVDGATR",
"AAAY(1)SAQVQPVDGATR", "", ""), `Uniprot ID` = c("Q9H7E9", "Q9H7E9",
"Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9",
"Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9", "Q9H7E9",
"Q9H7E9", "Q9H7E9", "Q00610", "Q00610"), `probe.Mod.or.not(Y/N)` = c("N",
"N", "N", "N", "N", "N", "N", "N", "N", "Y", "Y", "Y", "Y", "Y",
"Y", "Y", "Y", "Y", "N", "N"), `kinase.or.not(Y/N)` = c("N",
"N", "N", "N", "N", "N", "N", "N", "N", "N", "N", "N", "N", "N",
"N", "N", "N", "N", "N", "N"), Gene.Names = c("C8orf33", "C8orf33",
"C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33",
"C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33", "C8orf33",
"C8orf33", "C8orf33", "C8orf33", "C8orf33", "CLTC", "CLTC"),
Charge = c(2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 4L, 4L, 4L,
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L), m.z = c(802.90499, 802.90499,
802.90499, 802.90499, 802.90499, 802.90499, 802.90499, 802.90499,
802.90499, 647.57262, 647.57262, 647.57262, 647.57262, 647.57262,
647.57262, 647.57262, 647.57262, 647.57262, 1107.5251, 1107.5251
), Score = c(86.313, 86.313, 86.313, 86.313, 86.313, 86.313,
86.313, 86.313, 86.313, 41.695, 41.695, 41.695, 41.695, 41.695,
41.695, 41.695, 41.695, 41.695, 28.532, 28.532), Retention.time = c(27.774,
27.774, 27.774, 27.774, 27.774, 27.774, 27.774, 27.774, 27.774,
35.978, 35.978, 35.978, 35.978, 35.978, 35.978, 35.978, 35.978,
35.978, 72.556, 72.556), Precursor.Intensity = c(460631.45703125,
460631.45703125, 460631.45703125, 460631.45703125, 460631.45703125,
460631.45703125, 460631.45703125, 460631.45703125, 460631.45703125,
472201.625, 472201.625, 472201.625, 472201.625, 472201.625,
472201.625, 472201.625, 472201.625, 472201.625, 388790.9296875,
388790.9296875), Localization.prob = c(NaN, NaN, NaN, NaN,
NaN, NaN, NaN, NaN, NaN, 1, 1, 1, 1, 1, 1, 1, 1, 1, NaN,
NaN), probe_TMT6.site.IDs = c("", "", "", "", "", "", "",
"", "", "308", "308", "308", "308", "308", "308", "308",
"308", "308", "", ""), TMT_signals = c("TMT126", "TMT127N",
"TMT128N", "TMT128C", "TMT129N", "TMT129C", "TMT130N", "TMT130C",
"TMT131", "TMT126", "TMT127N", "TMT128N", "TMT128C", "TMT129N",
"TMT129C", "TMT130N", "TMT130C", "TMT131", "TMT126", "TMT127N"
), TMT_Intensities = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 1824.667,
3470.869, 1691.413, 2367.219, 1895.059, 1712.427, 1529.349,
1617.825, 1677.578, 0, 0)), row.names = c(NA, -20L), class = c("tbl_df",
"tbl", "data.frame"))
这是我从 ggplot 得到的图。
这是我从 Prism 获得的箱线图。
看起来 ggplot 和 Prism 计算的中值不同。例如，ggplot 中 126 Y 的中值约为 1500，而 Prism 中 126 Y 的中值约为 5000。
我检查了我的数据集，这些列中没有 NA 值。我自己计算的中值与 Prism 中的值相匹配。
有谁知道发生了什么事吗？
谢谢！
这里的问题出在 ylim 中，它与 scale_y_continuous(limits = ...) 一起有一种让一些用户感到意外的行为。如 ?ylim、
This is a shortcut for supplying the limits argument to the individual scales. By default, any values outside the limits specified are replaced with NA. Be warned that this will remove data outside the limits and this can produce unintended results. For changing x or y axis limits without dropping data observations, see coord_cartesian().
在汇总操作期间，这会产生特别令人困惑的结果，就像您在 geom_boxplot() 中得到的那样，因为它不会出错，它只会产生不同的结果和一个您可能会错过或忽略的警告.
例如，在下图中，我们期望箱线图的范围为 0 到 100，但只得到零值。这是因为使用 ylim 或 scale_y_continuous(limits = ...) 都会在执行任何汇总计算之前过滤掉范围外的数据。
df1 <- data.frame(x = 1, y = c(0, 100))
ggplot(df1, aes(x, y)) +
geom_boxplot() +
ylim(c(0,90))
有一个警告，但很容易错过或没有意识到它的含义：
Warning message: Removed 1 rows containing non-finite values (stat_boxplot).
这是说其中一个结果是 NA（在上述步骤之后）并且在统计计算中被忽略。
要得到预期的效果，一般应该使用coord_cartesian(ylim = ...)来控制观看window。
ggplot(df1, aes(x, y)) +
geom_boxplot() +
coord_cartesian(ylim = c(0,90))