Author Correction: Mapping QTLs conferring salt tolerance and micronutrient concentrations at seedling stage in wheat

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Growth conditions and phenotyping. The experiment was conducted in Venlo-type greenhouse located at 40°53′25″N, 29°22′47″E in Istanbul, equipped with computerized climate control for supplemental lighting, evaporative cooling and heating. The day and night temperatures were kept at 25 ± 4 °C and 20 ± 4 °C during the experiment period. 250 F 2 lines were germinated in perlite for 5 days and 180 seedlings were then transplanted into 2.7-L pots containing aerated nutrient solution 28 after removing their residual endosperm. On the following day, 75 mM NaCl was added to the nutrient solution. The solution was replaced every four days and the salinity level increased successively to 150, 225 and 300 mM NaCl on the 4 th , 8 th and 12 th day after transplantation. Plants remained under salinity for a total of 32 days including 20 days at 300 mM NaCl. Plants were then divided into four groups based on their phenotype: (i) tolerant (T) group had 5 fully extended green leaves without any signs of salt injury; (ii) moderately tolerant (MT) group had 4-5 fully extended green leaves with minor salt injury signs on leaf tips; (iii) susceptible (S) group had 2-3 leaves showing severe salt injury signs and 1-2 dead leaves; and (iv) highly susceptible (HS) group with 2-3 leaves having severe injury signs and death of 60-80% of leaves. Four representative plants from each group were selected for mineral analysis.
After three washings in dH 2 O, shoots and roots were dried at 65 °C for 3 days, dry root weight (DRW) and dry shoot weights (DSW) were recorded and tissues analyzed for mineral concentrations as previously described 28 . Dry roots and shoots were ground to powder in an agate vibrating cup mill (Pulverisette 9; Fritsch GmbH; Germany) and ~0.15-0.2 g powder from each sample was digested in 5 ml of 65% HNO 3 and 2 ml of 30% H 2 O 2 using closed-vessel microwave system (Mars Express; CEM Corp; NC, USA). Digested solutions were diluted with milli-Q water to 20 ml final volume for measurement of Ca, Cu, Fe, K, Mg, Mn, P, S, Zn and Boron concentrations by inductively coupled plasma optical emission spectrometry (ICP-OES; Vista-Pro Axial; Varian Pty Ltd; Mulgrave, Australia) 29 . The mineral concentration data was checked against standard values for standard reference material (SRM 8436 Durum Wheat Flour, NIST, Gaithersburg, MD). For measuring Na + concentration, solutions were further diluted 1:50. The data for 22 traits was obtained by multiplying ICP-OES values by the dilution factor and dividing the result by the DRW or DSW used for digestion. As reduction in the concentration of Na + means higher NAX so Na + values were multiplied by −1 to get root Na exclusion (RNAX) and shoot Na exclusion (SNAX) values. Correlation coefficients among the phenotypic data were calculated using Statistix 8.1 program. DNA extraction and genotyping. DNA extraction of 164 F 2 lines and the two parents was carried out using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA). DNA concentrations were measured with the Quant-iT PicoGreen dsDNA Assay Kit (ThermoFisher Scientific, Waltham, MA, USA) and 1.5 µg of gDNA from each line was dissolved in 10 mM Tris-HCl pH 8.0 to a final volume of 30 µl. Genotyping was performed using the Axiom Wheat Breeder's Genotyping Array (Affymetrix, Santa Clara, CA, USA) with 35,143 SNPs for each sample (hereafter referred to as the wheat 35 K array). Genotyping was done at the Bristol Genomics Facility (Bristol University, UK) utilizing the Affymetrix GeneTitan MT system following the Affymetrix procedure (Axiom 2.0 Assay Manual). We used Axiom Analysis Suite 1.1.0.616 software for SNP calling by following the Axiom Best Practices Genotyping Workflow with default wheat SNP call rate cut-off = 97%, QC call rate cut-off = 92% and DQC cut-off = 0.82. However, 10 lines failed the QC and DQC cut-offs; therefore, SNP call codes for the remaining 154 lines were used for downstream analysis. This software uses cluster separation, call rate and deviation from expected cluster positions and classifies the SNPs into six performance categories 19 .
Construction of SNP linkage map. The markers with significant segregation distortion (P < 0.05) were removed through chi-square test combined with sequential Bonferroni correction 30 . A genetic linkage map was constructed using MapDisto 2.0 b93 22 . In MapDisto, the markers were grouped with a recombination fraction of 0.3 and logarithm of the odds ratio (LOD) score of six using the Kosambi mapping function. The Seriation algorithm was used for ordering the linkage groups. Linkage groups were assigned to chromosomes by comparing shared markers with a recent wheat consensus linkage map 19 . Chromosomes were found to be divided into multiple linkage groups, so these linkage groups were combined and re-ordered. Rippling of marker order with a window size of five markers and checking for inversions were performed to improve the marker order and produce the shortest map of each chromosome. QTL mapping. Additive QTLs for all 22 traits were mapped by the composite interval mapping (CIM) method using single-treatment phenotypic data and QTL IciMapping V4.1.0 with 1-cM walking speed and LOD thresholds of 2.5 31 . The linkage maps and QTLs were drawn using MapChart 2.30 software 32 . The contribution of single QTL in phenotypic variation of mineral concentrations and salt tolerance was calculated by following Zhang et al. 33 . The DRW and DSW are considered to be the direct measure of salt tolerance 25,34 so the contribution of QTLs to salt tolerance was calculated by using DRW and DSW data.

Functional annotation of segregating SNPs. The flanking sequences of 3381 'Poly High Resolution'
(PHR) or polymorphic SNPs were mapped to gene coding sequences (CDS) from the International Wheat Genome Sequencing Consortium (IWGSC) database 35 using BLAST+ 2.2.30. From 1448 initial hits, basic local alignment search tool (BLAST) alignments with less than 60 bp length and/or 95% identity were removed. In this way, 1323 SNPs were found to be located in 1257 IWGSC CDS which were functionally annotated by using Blast2GO V4.0 36 . Functional annotation was done by using NCBI blast, mapping and annotation commands of Blast2Go using default parameters, and annotations with E-value ≥1 × 10 −30 were discarded.
In-silico expression analysis of annotated genes. For the purpose, above mentioned 1257 IWGS CDS were aligned with transcriptome reads expressed under normal and saline conditions 3 by using BLASTN 2.6.1+. Significant alignments with >200 alignment scores were counted and salt/normal alignment counts ratio was divided by 3.25 (347,200/106,600 spots for salt/control) to get the expression value. Genes with at least a 2-fold change in expression up or down were considered to be differentially regulated.

Results
SNP calling yielded six categories of SNPs. From 164 F 2 lines analysed on the wheat 35 K array, 154 gave good quality data and were used for SNP clustering. The SNPs were categorized into six groups: (i) PHR SNPs, which were polymorphic and co-dominant with a minimum of two samples containing the minor allele; (ii) Mono High Resolution (MHR) or monomorphic SNPs had only one cluster/allele; (iii) No Minor Homozygote (NMH), these polymorphic and dominant SNPs had only two clusters, one being the heterozygote; (iv) Off-Target Variants (OTV) had four clusters including one for a null allele; (v) Call Rate Below Threshold (CRBT) had all cluster properties above the threshold except for the call rate cut-off; and (vi) Other type SNPs, which had one or more cluster properties below quality thresholds (Fig. 1). Out of 35,143 array SNPs, 46.1% or 16,210 were MHR/ monomorphic and 51 (0.15%) SNPs were OTVs. Meanwhile 3,381 (9.6%) SNPs were PHR/polymorphic (Table 1) whose call codes were used for linkage map construction.
Whole genome wheat genetic linkage map. Out of 3,381 PHR markers, 1072 passed the chi-square test for segregation distortion and were used to construct the linkage map. We assigned 988 markers to 21 chromosomes and 84 markers were unlinked. The B genome had the highest number of markers assigned (562) followed by the A (342) and D (84) genomes. 183 markers were assigned to chromosome 1B while the fewest in the B genome (31) were assigned to 4B. For the A genome, maximum (100) and minimum (6) markers were assigned to chromosomes 3A and 6A respectively. In the D genome, chromosome 1D had 51 linked markers while 4D had only 2. The total length of the whole genome map was 2317.88 cM consisting of 975.56, 1133. 16  Comparison of linkage map with consensus map. By comparing our map with a recently published consensus map using these markers 19 , it was found that a total of 511 of 988 (51.7%) markers were assigned to the same chromosome as consensus map and 398 of 988 or 40.28% of the SNPs were mapped for the first time. The number of markers newly assigned to the A, B and D genomes was 132, 247 and 19 respectively. The maximum of these markers in the A genome i.e. 45 were located on chromosome 2A while 1B harboured 153 of these markers (Table S1). Remaining 79 markers were assigned to different chromosomes from the consensus map. The biggest such group consisted of 32 markers assigned to chromosome 2B in our map. While 17 such markers were assigned to chromosome 2A and 29 of total 79 SNPs were assigned to their respective homoeologous chromosomes (Table S2).  Table S3 and S4 respectively. We identified 49 additive QTLs for 20 single-treatment phenotypic traits on 12 wheat chromosomes. These QTLs were mapped on two chromosomes of the D genome and 5 chromosomes each from the A and B genomes ( Another RFeC QTL on 6B contributed 5.92% to salt tolerance. In the same region, a QTL for Mn contributed 14.16% of Mn phenotypic variation and 5.92% of salt tolerance. Although the QTLs for Boron, P and S contributed as much as 10.04% to mineral concentration variation, they made little contribution to salt tolerance (  (Table 4). Annotated sequences carrying polymorphic SNPs varied widely in function. For instance, 44 SNPs were annotated to ion transporters/channels linked to wheat salt tolerance, including seven potassium, two Chloride, four calcium, five zinc, four magnesium, two sulfate, three nitrate transporters/channels along with one proton pump, four proton transporters, cationic, anionic and metal channels ( Table 5, Table S5). Another 50 SNPs were annotated for protein, glucose, malate, drug, fatty acid and proline transport along with endocytosis/exocytosis (Table S5). Importantly, 92 SNPs were located on genes involved in Auxin, ABA and ethylene-activated and Jasmonic acid, gibberellic acid and sugar mediated signaling pathways; G-protein, apoptotic, cell surface receptor/Wnt signaling, signal recognition and transduction (Table S5). A further 63 SNPs were associated with genes from 35 different classes of TFs (Table S6). In total, 166 SNPs were located on genes with enzymatic activity (ligases, transferases, hydrolases), particularly kinases, which carried 77 SNPs. Some of these SNPs were closely related to post-translational modifications, such as phosphorylation and glycosylation, and to protein localization (Table S6). Genes involved in core cellular machineries, such as transcription, translation and replication, carried 198 SNPs (Table S7). Fifty-one SNPs were found within genes with Zn, Ca, Mg, Fe and metal ion binding activities, and 111 SNPs were associated with structural molecules, and thus, were part of different cellular organelles (Table S8). The highest number of SNPs (275) were confined to genes of various metabolic processes including photosynthesis, carbohydrate metabolism, oxidation-reduction, the Krebs cycle/respiration, protein catabolism, antioxidant activity, sugar, protein and lipid biosynthesis (Table S9).
Furthermore, 92 SNPs were associated with growth-related processes, including cell division/growth, meristem growth, autophagy, apoptosis, cell wall biogenesis/organization, lateral root development, leaf and shoot morphogenesis, flowering time; and flower, pollen, ovule and embryo sac development, while 48 more were located on genes related to stress responses, including disease resistance, drought, wounding, toxicity and freezing tolerance (Table S10). Finally, 109 SNPs were linked to hypothetical or uncharacterized proteins (Table S10).
In-silico expression analysis of candidate genes. Gene expression was categorised into six groups: up-regulated, expression not conclusive, down-regulated, expressed under salinity only, data not conclusive and no expression with 122, 425, 136, 156, 241 and 209 genes falling into these groups respectively (Table S11).

Discussion
The Wheat 35 K array is suitable for construction of genetic linkage map. We have constructed a high-density SNP linkage map of bread wheat using the wheat 35 K array, which consisted of probes for 35,143 exome-captured SNPs. Of these SNPs, 46.1% or 16,210 were monomorphic in our population and only 'PHR' SNPs (3,381 or 9.6%) were used for linkage map construction contrary to the recent paper 19 which used 'NMH' and 'OTV' SNPs as well. This was because we genotyped an F 2 population, so only PHR SNPs could be tested for typical F 2 segregation. Removal of markers with segregation distortion/bias is essential for getting a good quality linkage map, which was performed by Chi-square test with sequential Bonferroni correction 30 . From the 1072 SNPs passing the test, 988 were assigned to 21 chromosomes. Only 84 polymorphic markers were assigned to the D genome as compared to 342 and 562 markers assigned to A and B genome. This has been linked to lower nucleotide diversity of the D genome due to its relatively recent evolutionary origin 37,38 . Our genetic map has a total length of 2317.88 cM in comparison to 3739.23 cM for the published wheat consensus map 19 , largely due to the lack of diversity/segregation in the D genome under salinity. Of 988 SNPs, 398 SNPs were assigned to chromosomes for the first time, and 79 markers were assigned to different chromosomes from the consensus map. These novel and conflicting markers show genetic diversity originating from genetic differences between European and Pakistani wheat lines. Secondly, we used an F 2 population instead of inbred lines, which may have contributed in different segregation patterns. Even so, the large majority (511 of 590) of markers common to both maps were assigned to the same chromosomes 19 .
Several novel and major QTLs were mapped for minerals and salt tolerance. We mapped 49 QTLs for 22 traits under salt stress on 12 chromosomes, including four QTLs on two chromosomes of the D genome. Salt tolerance in wheat is mainly conferred by NAX or reduced Na + uptake, as Na + influx leads to reduce photosynthesis, growth, development and yield 5,7,9 . The genetic basis of NAX was unknown until the identification of a major NAX QTL on chromosome 2A 39 ; and QTL mapping for salt tolerance has focused on NAX in recent years 23,25,26 . Among the six mapped RNAX and SNAX QTLs, two closely located QTLs (qSNAX.2A.1 and qSNAX.2A.2) and qRNAX.2A.1 coincided with the major NAX locus Nax1 or HKT1;4 in durum wheat 39 and three NAX QTLs found on 2A in bread wheat 25 . Similarly, the qRNAX.6A.2 QTL has also been reported previously 25,26 . We mapped two novel and major QTLs (qSNAX.7A.3 and qRNAX.7A.3) on chromosome 7A, which contributed 13.69 and 15.35% of the phenotypic variation of RNAX and SNAX and accounted for 11.23 and 19.79% of the observed salt tolerance (DRW and DSW respectively). The HKT genes are well characterized regulators of K + and/or Na + transport in plants and encode proteins that reduce Na + transport to the shoot, thus conferring salt tolerance 5 . Accordingly, QTL mapping for K + conc. under salinity has also been performed by some groups 23,25-27 . We identified a major SKC QTL on 6A contributing 7.46 and 9.76% of the phenotypic variation in K + conc. and salt tolerance. Another novel and major QTL (qRKC.4B.2) presented 11.31% of phenotypic variation of RKC. Another RKC QTL mapped on 3D was consistent with a reported QTL 25 and the remaining SKC and RKC QTLs were co-located with the SNAX and RNAX QTLs. This co-localization of QTLs is explained by the functional correlation between these traits.  Due to the focus on NAX and K + QTLs, to our knowledge the genetics of Ca 2+ and Mg 2+ accumulation under salinity has only been investigatedonce 27 . We mapped two major QTLs for RCalC (qRCalC.6B.3 and qSCalC.6B.2), which contributed 5.92 and 11.87% of salt tolerance and 10.91 and 6.52% of Ca phenotypic variation. The above mentioned and two RCalC QTLs on 3B coincided with previously reported QTLs 27 , but a novel QTL on 6A also presented 8.98% of the observed SCalC phenotypic variance. Among Mg 2+ (RMgC and SMgC) QTLs, a 2A QTL presented 6.37% of SMgC phenotypic variation and qSMgC.6B.2 contributed 8.36% to salt tolerance. Another novel QTL on 5A accounted for 5.58% of salt tolerance. We also mapped 27 novel QTLs for P, Zn, Fe, Mn, Cu, S and Boron concentrations in wheat root and shoot under salinity for the first time as QTLs for these mineral were previously reported under different water regimes only 24 . Among them, the most important major QTL for RZnC on 7A contributed 11.23 and 12.08% to salt tolerance and RZnC phenotypic variation. One each of the FeC and MnC QTLs, found on 6B, presented 5.92% of salt tolerance and the Mn QTL contributed 14.16% to Mn phenotypic variation. Similarly, a major qRFeC.6A.2 contributed 12.96% of Fe phenotypic variation. Despite the low contribution of Boron, P and S QTLs to salt tolerance, their contribution to mineral phenotypic variation under salinity was high, which could be useful information for future wheat breeding.
We found five QTL clusters on the A and B genome containing QTLs for several minerals. Such clusters are expected as changes in cellular Na + could affect concentrations of many other ions; indeed, previous studies have shown clustering of QTLs for closely correlated salt traits 23,24,26 . Each cluster may represent either a single gene or several closely linked genes; e.g. the occurrence of two closely linked NAX QTLs on 2A and the co-localization of Zn and NAX QTLs. This co-localization or QTL clustering is due to high the correlation coefficient among the traits of the respective QTLs as highlighted (Table S4).

Functional annotation of segregating SNPs highlighted complexity of salt tolerance.
Segregating F2 populations are valuable material for the dissection of the genetic architecture of complex traits, as they depict maximum segregation in phenotype and polymorphism at molecular level 14,39 . The Wheat 35 K Array contains exome-captured SNPs; thus, flanking sequences of segregating SNPs could be used for functional annotation, which revealed that genes of various biological processes and molecular functions are likely to be involved in salt tolerance mechanisms. SNP-carrying sequences were highly similar to those from A. tauschii and T. urartu, reflecting their close ancestral relationships with bread wheat 37 . Another top species from BLAST comparisons was H. vulgare, the most salt tolerant cereal 15 ; indicting shared tolerance genes between wheat and barley. We annotated 44 ion transporters responsible for ion homeostasis. Among the seven annotated K transporters/channels, the role of HKTs i.e. TmHKT1;5-A Nax2, HKT1;4 Nax1 as sodium excluders 5,25 , K + /H + exchanger, TaNHX2, as K + /H + antiporter 40 and Jacalin-related lectin 3 (TaJRL3) in salt tolerance response 41 was reported previously. The K + outward-rectifying channel (KORC) negatively regulates salt tolerance by K + efflux from plant roots under salinity 42 , but the functions of the remaining K + channels are not known. The annotated chloride channels (CLC-g and CLC-e) could be linked to previously mapped Cl − QTLs including a major 5A QTL 27 . The annotated Pyrophosphate-energized proton pump (H + -PPase or TVP1) and four proton transporters were previously found to induce the sequestration of Na + ions into the vacuole and act as Na+/H+ antiporters to confer salt tolerance in wheat 43 . However, none of the four Ca, five Zn transporters, three nitrate transporters and two each of Mg, Sulfate and Cu-transporters have been characterized for their salt tolerance response in wheat. Another 50 SNPs annotated for transport of proteins, mRNA, glucose, malate, fatty acids and proline need further investigation under salinity.
Understanding of complex abiotic stress signalling pathways is essential for a successful breeding program 34,44 . In total 92 SNPs were located within genes involved in several signalling pathways, e.g. signal recognition, transduction, cell surface receptor and Wnt signalling to identify stress stimuli. Annotation of apoptotic signalling genes could be linked to the hypersensitive response to avoid stress injury 45 . Similarly, a total of 14 genes were identified for ABA and ethylene activated signalling and for JA and SA-mediated signalling pathways, which are thought to be involved in conferring salt tolerance 3 . Moreover, some genes for ABA, ethylene and JA signalling also confer salt tolerance in wheat 46,47 . We identified 22 genes for auxin biosynthesis, transport and auxin-activated signalling pathways, but the role of auxin signalling under salinity is not known. TFs are proteins that regulate the expression of several stress-related genes 34 and 63 annotated SNPs were linked to 35 classes of 52 TFs e.g. MYB44, WRKY16, WRKY70, bZIP17, NAC17, NAC78 etc. However, bHLH140, GATA26, ZNFX1-NFXL1, EIN3, ABI3, ARF3, ARF5 and ARF21 appear to be major salinity responsive TFs based on their GO annotations. Only the TFs ABI3 and ARF3 have previously been reported to be involved in salt tolerance mechanisms and signalling 46,48 in wheat. Interestingly, 198 SNPs were located on genes involved in various nucleic acid processes including chromatin modifications, helicases, DNA repair mechanisms, DNA replication, transcription and translation regulation. Some of these genes were over-expressed under salinity 3 . Other such gene groups coded for epigenetic (DNA, tRNA, rRNA and histone-lysine methylation) or epi-transcriptomic (splicosomal complexes, splicing site recognition, mRNA splicing, mRNA based gene silencing) processes. As these reactions are core cellular functions  they may regulate the expression of salinity related genes 49 . Additionally, 166 SNPs were located on genes with enzymatic (ligases, transferases, hydrolases, isomerase and kinase) or protein modification activity. These included 77 kinases, protein kinases and protein serine/threonine kinases coding for protein phosphorylation or post-translational protein modification. Protein serine/threonine kinases and protein kinases play a role in wheat salt tolerance and ABA signaling respectively 48,50 . Genes for other post-translational processes (protein de-phosphorylation, dimerization and glycosylation) were also identified. Another 51 SNPs were linked to genes involved in Zn, Ca, Mg, Fe and metal ion binding, as metal ions are part of several enzymes that may have direct roles in conferring salt tolerance. The largest number of annotated SNPs, 275, were linked to genes of metabolic processes. The biggest subgroup consisted of 65 SNPs associated with photosynthesis regulation, chloroplast fission/organization, chlorophyll biosynthesis/catabolism, Photosystem (PS) I and II complex, PS II assembly, light reaction, Carbon/ energy pathway, chloroplast DNA synthesis/translation; and synthesis of photosynthetic sugars (glucose, galactose, fructose, mannose). The role of these genes in photosynthesis under salinity in wheat has not been thoroughly investigated, but 22 photosynthetic proteins showed differential expression under salinity in wheat 51 . Additionally, 45 genes were identified for oxidation-reduction (Redox) processes, which are the backbone of cell mechanisms. A redox gene, 12-oxophytodienoate reductase 1 (OPR1), confers salt tolerance to wheat by enhanced ABA signalling and reactive oxygen species (ROS) scavenging 52 . The redox genes coding for peroxidase 1, 2 and 12, and 19 genes for synthesis of antioxidants (glutathione, flavonoid, carboxylic acid, lactate, cytokinin, vitamin B and E) could be involved in ROS scavenging under salinity induced osmotic stress. Genes for ubiquitin, proteasome and proteolysis dependent protein catabolism could destroy the unwanted proteins under salinity. However, roles of 21 respiratory genes (photorespiration, respiratory chain complex I and II, glycolysis, and Kreb cycle) and 64 genes for biosynthesis of carbohydrates, lipids and proteins needs to be investigated. Among the 92 SNPs associated with genes of several growth stages such as root/shoot development to reproductive growth (flowering time, pollen germination, ovule development etc.), inheritance of flowering time under salinity has been reported 15 .
Several annotated genes were transcriptionally expressed. In order to validate the role of annotated SNPS/genes in conferring salt tolerance, they were aligned with published transcriptome data describing genes that are differentially expressed under salinity 3 . A total of 122 of the annotated genes were up-regulated in silico, which included protein, mRNA, nitrate and cations transporters; several Wnt/ABA/auxin activated signaling molecules; signal transduction, kinases, proteolysis, REDOX process, flavonoid metabolism, defense response, cell wall/xylum development genes etc. On the other hand, 136 genes were down-regulated under salinity. The only available salt-expressed transcriptome in wheat root is based on a single genotype; therefore, several of the annotated genes, such as photosynthetic genes, were not expected to show differential expression.

Conclusions and prospects.
We have identified two novel major QTLs on wheat chromosome 7A which contributed 11.23 and 15.79% to salt tolerance, and 13.69-15.35% to NAX in our population. Another major Zn QTL contributed 12.08 and 11.23% to Zn phenotypic variation and salt tolerance respectively. A major SCaIC QTL also accounted for 11.87% of the observed salt tolerance trait. We have also identified other novel QTLs which contributed 10.91, 12.96,11.31 and 14.16% of Ca, Fe, K and Mn phenotypic variation. Other mapped QTLs represented 2.1-8.98% of phenotypic variation of different minerals. These novel QTLs could be used for could be used for MAS breeding for salt tolerance and breeding for biofortification of wheat for Ca, Zn, Mg, Fe and Mn. We have also annotated 1293 segregating SNPs, which were located within genes for various ion channels, signalling pathways, TFs, metabolic pathways etc. and 258 of them were differentially expressed under salinity, indicating that they may have a role in salt tolerance. As the published transcriptome used for this analysis is based on Roche 454-GS FLX sequencing, future transcriptome data fromsalt stressed wheat using Illumina technology could help to further understand gene expression under salinity. The characterization of the annotated genes described here will help to dissect salt tolerance mechanisms, guiding future breeding for this important trait.