Table S1. Representative phenotypic and agricultural traits of the red-purple (var. “Eumjinara”) and yellow (var. “Sinsunhwang”) onion. All numbers indicate average value for each trait.
Traits/Variety / Eumjinara / SinsunhwangProduction (kg/10a) / 6,999 / 5,929
Sweetness (Brix)A / 7.67 / 5.6
Leaf type / Open / Open
Waxy / Medium / Medium
Leaf color / Green / Green
Bulb shape and color / Spherical, red-purple / Big globular, yellow
Bulb weight (g) / 234 / 254
Plant height (cm) / 62 / 60
No. leaves / 7.7 / 7.6
Resistance / Cold / Strong / Medium
Downy mildew / Somewhat strong / Medium
Soft rot / Medium / Medium
ASugar contents of aqueous solution. One degree Brix is one gram of sucrose in 100 grams of solution, represents the strength of the solution as percentage by mass
Table S2. Short-read RNA sequencing generated by Illumina HiSeq 2500 for DEG analysis with three replications. The short-reads were mapped to the reference transcripts.
Tissue(Replicates) / Read order / IndexA / Yield (bases)B / # ReadsC / % of ≥Q30 bases (PF)D / Mean quality score (PF)
Bulb-1 / 1 / CGATGT / 2,424,111,302 / 24,001,102 / 92.08 / 35.93
2 / 89.06 / 35.33
Bulb-2 / 1 / GGCTAC / 2,353,163,549 / 23,298,649 / 91.23 / 35.75
2 / 89.42 / 35.40
Bulb-3 / 1 / ACAGTG / 2,805,622,339 / 27,778,439 / 91.84 / 35.88
2 / 89.09 / 35.36
Root-1 / 1 / TAGCTT / 2,386,586,571 / 22,639,471 / 91.61 / 35.83
2 / 90.09 / 35.54
Root-2 / 1 / TGACCA / 2,668,892,781 / 26,424,681 / 91.08 / 35.71
2 / 90.00 / 35.53
Root-3 / 1 / ATCACG / 2,847,840,440 / 28,196,440 / 91.09 / 35.71
2 / 88.55 / 35.24
Leaf-1 / 1 / GCCAAT / 2,765,195,473 / 27,378,173 / 91.21 / 35.73
2 / 89.81 / 35.47
Leaf-2 / 1 / TTAGGC / 2,513,766,477 / 24,888,777 / 91.53 / 35.81
2 / 89.80 / 35.47
Leaf-3 / 1 / GATCAG / 2,164,333,040 / 21,429,040 / 91.99 / 35.91
2 / 88.49 / 35.24
Flower-1 / 1 / CAGATC / 3,244,678,227 / 32,125,527 / 91.64 / 35.81
2 / 88.97 / 35.30
Flower-2 / 1 / ACTTGA / 3,197,176,513 / 31,655,213 / 91.54 / 35.80
2 / 89.80 / 35.47
Flower-3 / 1 / CTTGTA / 2,516,935,251 / 24,920,151 / 92.05 / 35.93
2 / 88.80 / 35.28
A Adaptor sequence for library construction
B Number of bases sequenced
C Number of fragments sequenced with 101 bp
D Threshold for sequence quality
Table S3. The ratios of obtaining polished high-quality isoforms (PHQIs) from the total sequence reads of 8 SMRT cells through Consensus calling (ICE) and Quality filtering (Quiver). The PHQIs were generated using IsoSeqTM (SMRT® Portal version 2.2.0).
Procedures / Classification / No. reads(isoforms) / Ratio(%)ASequence reads / Total No. reads / 511,761 / 100
No. 3’-prime reads / 238,652
No. 5’- prime reads / 238,058
No. poly(A) reads / 232,102
Structure / No. filtered short-reads / 15,948 / 3
No. non-full-length reads / 360,752 / 70
No. full-length readsB / 135,061 / 26
Consensus calling (ICE) / No. full-length non-chimeric reads / 134,579 / 25.9
Average length of full-length non-chimeric read / 2,454
No. consensus isoformsC / 63,737 / 12
Average length of consensus isoforms / 2,610
Quality Filtering (Quiver) / No. polished high-quality isoformsD / 5,786 / 1
No. polishedlow-quality isoforms / 5,989
A Ratio of read numbers to total No. reads
B No. reads containing 5’-, 3’- and poly(A) reads altogether
CNo. reads containing PHQI, PLQIs
D No. misread-corrected polished reads with consensus isoforms
Table S4. The comparison process of obtaining polished high-quality isoforms (PHQIs) from the total reads between 8 and 16 SMRT cells. The PHQIs were generated using IsoSeqTM (SMRT® Portal version 2.2.0)
Structure/ processing / Classification / No. reads (isoforms)8 cells / 16 cells
Consensus calling (ICE) / No. consensus isoforms / 63,737 / 113,316
Average length of consensus isoforms / 2,610 / 2,645
Quality Filtering (Quiver) / No. polished high-quality isoforms / 5,786 / 10,449
No. polished low-quality isoforms / 5,989 / 8,800
Table S5. Production of PHQIs and average length from each tissue: flower, leaf, bulb and root. Due to overlapped gene expression among the tissues, total reads from the four tissues were combined to produce proper PHQIs. The PHQIs were generated using an upgrade version of IsoSeqTM (SMRT® Portal version 2.3.0).
Tissue / No. cellA / Consensus isoforms / PHQI / PLQIB / Average length (bp)Flower / 16 / 108,508 / 20,956 / 87,552 / 2,290
Bulb / 16 / 139,929 / 21,627 / 118,302 / 2,377
Leaf / 16 / 131,921 / 22,532 / 109,389 / 2,133
Root / 16 / 140,513 / 23,353 / 117,159 / 2,264
TotalC / 64 / 520,871 / 88,468 / 432,402 / 2,266
Combined analysisD / 64 / 482,335 / 99,247 / 383,086 / 2,251
A No. of SMRT cells sequenced
B No. polished low-quality isoforms (PLQI)
C Total sum of numbers (cells, isoforms, PHQI, PLQI and average length) from four tissues
DNo. consensus isoforms, PHQI, PLQIs and average length through combined analysis with total reads from four tissues
Table S6.The ratios of full-length cPHQIs randomly-selected from the 35,505 cPHQIs based on size. The full-length structures contain both a translation start (ATG) and stop (TAG/TAA) codon, identified using BLAST(x/n).
Ref. transcript(kb) / No. ref. transcript / No. unknownB / Sub-totalCFull-length / Partial-lengthA
0 to1 / 13 / 4 / 5 / 22
1 to 2 / 107 / 8 / 43 / 158
2 to 3 / 65 / 5 / 37 / 107
3 to 4 / 34 / 3 / 15 / 52
4 to 8 / 8 / 1 / 4 / 13
Total (%)D / 227 (64) / 21(6) / 104 (30) / 352 (100)
A No. cPHQIs possessing either start orcodon, or neither
B No. cPHQIs with unknown function
CSub-total of number of cPHQIs by size
D No. (ratios) of cPHQIspossessing full-length structure
Table S7. Comparison of completeness between 35,505 cPHQIs and assembled contigs. The contigswere assembled with 629,471,326 short-reads (HiSeq 2500, 101×101 bp) using Trinity(Grabherr et al. 2011)
Sizes (kb) / cPHQIs / Assembled contigsNo. cPHQIs / Ratio (%) / No. contigs / Ratio (%)
Less than 1 / 2,152 / 6 / 672,607 / 82
1 to 2 / 15,766 / 44 / 103,571 / 13
2 to 3 / 10,727 / 30 / 28,412 / 3.5
3 to 4 / 5,296 / 15 / 8,632 / 1
4 to 5 / 1,561 / 4.7 / 2,840 / 0.3
5 to 6 / 3 / 0.1 / 1,030 / 0.1
6 to 7 / 351 / - A
over 7 / 387 / -
No. isoforms/contigs / 35,505 / 817,830
Longest size (bp) / 7,943 / 19,228
Shortest size (bp) / 267 / 201
Mean size (bp) / 2,235 / 669
Avalue too small
Table S8. The essential information of tissue-specific isoforms of root, bulb, leaf and flower as shown in Fig. 1. The tissue-specific genes were analyzed by DESeq R-package. The cut-off values of tissue-specific expression and adjusted P-value (padj) are 3-fold and ≤ 0.05, respectively.The information consists of ID, length (bp), accession (NCBI GenBank), putative gene function and RPKM value of each tissue-specific isoforms. No hit and unknown indicate that there is no registered sequence in NCBI (GenBank) sharing a high homology with theisoform[Sohn1].
1
[Sohn1]According to major point 1, the list of 4 tissue-specific isoforms are attached as an Excel file.