From By-Products to Fertilizer: Chemical Characterization Using UPLC-QToF-MS via Suspect and Non-Target Screening Strategies

The increasing demands of agriculture and the food market have resulted in intensive agricultural practices using synthetic fertilizers to maximize production. However, significant efforts have been made to implement more environmentally friendly procedures, such as composting, to overcome the adverse impact of these invasive practices. In the terms of this research, composting was applied to the production of two biofertilizers, using onion and mushroom by-products as raw materials respectively. The main purposes of this work were to identify the compounds that pass from the raw materials to the final products (onion-based and mushroom-based), as well as the characterization of the chemical profile of these final products following suspect and non-target screening workflows via UPLC-qToF-MS. Overall, 14 common compounds were identified in the onion and its final product, while 12 compounds were found in the mushroom and its corresponding product. These compounds belong to fatty acids, organic acids, and flavonoids, which could be beneficial to plant health. The determination of parameters, such as the pH, conductivity, organic matter, nitrogen content, and elemental analysis, were conducted for the overall characterization of the aforementioned products.


Introduction
The demands of the agri-food market in contemporary society have resulted in the global and national proliferation of intensive agricultural practices. In this direction, the use of synthetic fertilizers has increased in order to maximize their production. However, chemical fertilizer utilization that exceeds a certain threshold level pollutes water bodies in addition to causing accumulation in crop plants [1]. Therefore, considerable effort has been expended to seek more environmentally friendly methods of increasing productivity while also ensuring agricultural biosafety with no adverse effects on soil microflora, the environment, or human health.
Composting is the aerobic decomposition of organic materials into a humus-like material through microbial action [2]. Research found that soil supplementation with compost acting as fertilizer is beneficial towards improving the physical chemical properties and nutrient status [3]. Furthermore, compost is a natural and environmentally friendly product that is well known for its beneficial effects on soil and plant nutrition, whereas, due to the recycling of organic waste, compost can be considered as either an alternative [4] or a fertilizer booster [5]. The utilization of compost as a fertilizer is also in accordance with the EU guidelines for the reduction of non-renewable sources of fertilizer up to 30% (priority axis for Europe 2020) [6].
Thus, the recycling of fertilizer components through composting, as well as the diminished environmental impact due to lowering transportation needs, render this approach as

Results
Two final products (onion-based and mushroom-based) were derived after the hydrolysis of the compost from the two raw materials. The results derived through the suspect and non-target screening workflows are described in the following sections.

Suspect and Non-Target Screening of the Raw Onion Material and the Onion-Based Final Product
A literature-based suspect list for the onion-based final product was assembled, comprising natural products appearing in onions. A total of eight compounds were identified from this list, both in the raw material and the onion-based final product. Six compounds (myristic acid, palmitic acid, linoleic acid, oleic acid, quercetin, and citric acid) reached identification confidence level 1, whereas two compounds (isorhamnetin-4-glucoside and fumaric acid) were identified at the 2a confidence level.
In total, six compounds were identified through the non-target screening workflow. Three compounds (C17-sphinganine, glycerin palmitate, and monostearin) were identified at the confidence level 2a, two compounds (isosakuranetin and x,y-dihydroxybenzaldehyde) reached identification level 3 and one compound reached identification level 5.
The compounds identified by suspect and non-target screening both in the raw material and the final product are tabulated in Table 1 along with their molecular formula, the experimental and predicted retention time, the theoretical and experimental m/z value of the precursor ion and the ionization mode, as well as the five most abundant MS/MS fragments (if they existed) of the raw material and final product in comparison to the MS/MS fragments of the corresponding spectrum in the spectral library or reference standard.

Suspect Screening and Non-Target Screening of the Raw Mushroom Straw Material and the Mushroom-Based Final Product
Ten compounds were identified from a suspect list compiled using scientific literature for mushrooms and were identified in both the raw material (mushroom straw) and the corresponding final product. Six compounds (myristic acid, palmitic acid, linoleic acid, oleic acid, stearic acid, and leucine) reached the highest identification level (level 1). Furthermore, four compounds (4-hydroxybenzaldehyde, choline, agmatine, and spermidine) were identified at the 2a level of confidence. Moreover, the identification of C17-sphinganine (confidence level 2a) and x,y-dihydroxybenzaldehyde (identification level 3) derived from onions, were components of the final product based on mushroom straw.
Two compounds (9-hydroxy-8,10-dehydrothymol, x,y-dihydroxy-benzaldehyde x = 3, 2, 2, 2 y = 4, 4, 5, 3) were identified through the non-target screening workflow, reaching confidence level 3. Table 2 illustrates the compounds identified by suspect and non-target screening in both the raw material and the final product. This table also includes their molecular formula, experimental and predicted retention time, theoretical and experimental m/z value of the precursor ion and ionization mode, and the five most-abundant MS/MS fragments (if they existed) of the raw material and final product in comparison to MS/MS fragments of the corresponding spectrum of the spectral library or reference standard.

Supplementary Chemical Analysis
The results of the physicochemical (conductivity, pH) as well as elemental analysis are summarized in Appendix A.1 and in Table S1 (Supplementary Materials).

Suspect Screening of the Raw Onion Material and the Onion-Based Final Product
A literature-based suspect list for the onion-based final product was assembled, comprising natural products appearing in onions. [13][14][15][16][17][18][19]21,[33][34][35][36]. A total of eight compounds were identified from this list, both in the raw material and the onion-based final product. Six compounds (myristic acid, palmitic acid, linoleic acid, oleic acid, quercetin, and citric acid) reached identification confidence level 1, whereas two compounds (isorhamnetin-4glucoside and fumaric acid) were identified at the 2a identification confidence level.
Quercetin was used as an example of the methodology of the suspect screening. The theoretical m/z value of 301.0354 was calculated using Bruker's Isotope Pattern software, version 4.3 (Bruker Daltonics, Bremen, Germany), and the extracted chromatographic peak for that m/z value was observed at 7.1 min for both the raw material and the final product. The extracted ion chromatograms of the 4 and 25 eV collision energy were compared concluding that this m/z value is not a produced ion. In addition, no chromatographic peak for this m/z value was observed in the blank sample ( Figure 1a). Furthermore, the five most abundant ions (higher MW ions than the ion under consideration) for this precursor were investigated to ensure that it is not an in-source fragment. Thus, the m/z value 301.0354 (±0.005 Da) is not derived from any other ion. Although the chemical structure of the compound is known in the case of suspect screening, the "Smart Formula Manually" software (Bruker Daltonics, Bremen, Germany) was used to evaluate mass accuracy and assign the molecular formula. The neutral molecular formula of C 15 H 10 O 7 was found for the raw material and the final product was assigned with mass accuracy (2.0 ppm/ 0.6 mDa, 1.5 ppm/0.4 mDa) and isotope fitting (8.7 mSigma and 40.5 mSigma-a value < 100 mSigma is considered confirmatory of the isotopic fitting), respectively. Furthermore, the compound's elution time in the raw material and final product (Exp. t R = 7.1 min) was within the 1.8 min confidence margin of the predicted value (t Rpred = 7.0 min) [32]. The Monte Carlo sampling (MCS) method [37] was used to evaluate the difference between the experimental and theoretically predicted time and to eliminate candidates that are false-positive hits (Figure 1b) (can be accessed at the available online tool "Retention Time Indices Platform (RTI platform)" http://rti.chem.uoa.gr/ (accessed on 8 March 2021)). Next, the MS/MS fragments with a signal-to-noise ratio less than 3 were excluded from the MS/MS spectrum list obtained for m/z 301.0354, and the remaining ones were normalized to be compatible with the MetFrag. This chemical structure corresponds to quercetin and its MS/MS fragments were loaded to MetFrag. All three fragments (107.0143, 151.0036, and 178.9980) were explained by the algorithm. The fragments and their corresponding chemical structures are illustrated in Figure 1c. The MS/MS spectrum of quercetin was compared to the one available in MassBank (MassBank ID: FIO00279). The MS/MS fragments (151.0035 and 178.9986) of the library entry are matched with the experimentally derived fragments (151.0036 and 178.9980) of the suspect compound quercetin in the final product. Following the confirmation of the fragments with the corresponding MS/MS spectra in the spectral library (MS/MS similarity score of 0.961 and 0.870 for the raw material and final product, respectively) and the agreement of the experimental retention time with the predicted one, the confidence level of identification for quercetin assigned to 2a. The common fragments of the final product and the standard spectrum (MassBank ID: FIO00279) are depicted in Figure 1d. Finally, the sample's MS/MS spectra and elution time were compared to a commercially available standard (t R = 7.1 min). Figure 1e shows the common MS/MS fragments of the product and those of the standard (MS/MS similarity score of 0.961 and 0.922 for the raw material and the final product, respectively). As a result, quercetin was assigned to identification level 1.
the standard (MS/MS similarity score of 0.961 and 0.922 for the raw material and the final product, respectively). As a result, quercetin was assigned to identification level 1. For the theoretical m/z value 227.2017, a chromatographic peak at 13.0 min (raw material and final product) was observed, corresponding to the candidate myristic acid from the suspect list. The neutral molecular formula C14H28O2 was assigned to this mass (mass accuracy of 0.4 ppm/0.1 mDa and 0 ppm/0 mDa) isotopic fit (5.2 and 13.6 mSigma) for the raw material and the final product. The compound was identified using MassBank ID: For the theoretical m/z value 227.2017, a chromatographic peak at 13.0 min (raw material and final product) was observed, corresponding to the candidate myristic acid from the suspect list. The neutral molecular formula C 14 H 28 O 2 was assigned to this mass (mass accuracy of 0.4 ppm/0.1 mDa and 0 ppm/0 mDa) isotopic fit (5.2 and 13.6 mSigma) for the raw material and the final product. The compound was identified using MassBank ID: BS003878 with a level of confidence of 2a. After comparing its MS/MS spectra and retention time with the reference standard (t R = 12.9 min), the identification of myristic acid was confirmed. As a result, myristic acid was assigned the identification level of 1.
For the theoretical m/z value 255.2330, a chromatographic peak eluting at 13.8 min for both raw material and final product was detected, corresponding to the candidate palmitic acid. This mass was assigned to a neutral molecular formula of C 16 H 32 O 2 with mass accuracy of 0.7 ppm/0.2 mDa and 1.0 ppm/0.3 mDa and an isotopic fit of (4.2 mSigma and 7.6 mSigma). Moreover, the predicted retention time (t Rpred = 12.9 min) was consistent with the corresponding experimental retention time. The MS/MS fragments matched those in the MoNA spectral Library (MoNA ID: MetaboBASE1104), reaching a confidence level of 2a. The samples' MS/MS spectra and retention times were then compared to those of the reference standard (t R of standard = 13.8 min), and they were found to be in agreement. As a result, palmitic acid reached a confidence level of 1.
A compound (linoleic acid) from the suspect list was matched for the m/z value 279.2330, eluting at 13.4 for both the raw material and final product. For this m/z value, both samples met all of the identification criteria (29.6 and 23.9 mSigma; isotopic fit) and (−0.7 ppm/−0.2 mDa and 0.3 ppm/0.1 mDa; mass accuracy) for the neutral molecular formula of C 18 H 32 O 2 . Furthermore, the predicted retention time corresponded to the experimental results. The MS/MS fragments were evaluated by comparing them to the corresponding reference spectrum in the MassBank library (MassBank ID: RP029511). The MS/MS spectrum and retention time of the reference standard (t R = 13.5 min) were then compared to the samples. Therefore, linoleic acid reached a level of confidence of 1.
The m/z value 281.2486 is detected in both samples (t R = 13.9 min) and corresponds to oleic acid, which was a candidate in the suspect list. The neutral molecular formula of C 18 H 34 O 2 was assigned to this mass with mass accuracy of 1.3 ppm/0.4 mDa, isotopic fit of 8.4 and 31.3 mSigma). It is assigned at the level of identification 2a by comparing fragments from the MS/MS spectrum of the reference standard with the attributed in the mass spectral library (MassBank ID: BS003989). The reference standard's MS/MS spectrum and retention time (t R = 14.0 min) are in close agreement to those of the samples, reaching the level of confidence 1.
A chromatographic peak was observed eluting at 1.1 min (raw material and final product) for the m/z value 191.0197, which could be attributed to citric acid from the suspect list with assigned neutral molecular formula C 6 H 8 O 7 (mass accuracy of −7.1 ppm/−1.4 mDa, isotopic fit of 16.9 mSigma) calculated for the feature of the raw material. The identification of citric acid is confirmed after comparing its MS/MS spectra to the corresponding MassBank spectrum with ID: PR100481, as well as to the reference standard. Additionally, the retention time of the reference standard (t R = 1.1 min) was found to be identical to the retention time of the suspect compound in the samples. As a result, citric acid reached level 1.
The theoretical value m/z 477.1039 corresponds to the suspect compound isorhamnetin-4-glucoside, with elution time of 6.5 min for both compounds (raw material and final product, respectively). This mass was assigned to the neutral molecular formula C 22 H 22 O 12 (mass accuracy of 1.8 ppm/0.9 mDa and −0.4 ppm/−0.2 mDa; isotopic fit (28.1 mSigma and 7.7 mSigma) for the raw and final product. The predicted retention time (t Rpred = 6.5 min) is consistent with the experimental retention time of the samples. The MS/MS spectrum with MassBank ID: PR040093 was used to identify this compound, yielding a level of confidence of 2a.
The theoretical m/z value 115.0037, eluting at 1.1 min (raw and final product), corresponds to the candidate fumaric acid from the suspect list. For the raw material, the neutral molecular formula C 4 H 4 O 4 was assigned (mass accuracy of 0.8 ppm/0.1 mDa, isotopic fit (52.4 mSigma). The predicted retention time (t Rpred = 2.0 min) is in good accordance to the samples' experimental retention time. After MS/MS evaluation with the corresponding spectrum of MzCloud (No ID: 1274), the compound was identified at the 2a confidence level.

Non-Target Screening of the Onion and Its Corresponding Final Product
In total, six compounds were identified through the non-target screening workflow. Three compounds (C17-sphinganine, glucerin palmitate, and monostearin) were identified at the confidence level 2a, two compounds (isosakuranetin and x,y-dihydroxybenzaldehyde) reached identification level 3, and one compound reached identification level 5. The no-tation followed for the mass spectral data corresponds to the experimental m/z value raw material /m/z value final product .
The m/z 285.0778/285.0767 chromatographic peak eluted at 9.2 min for both the raw material and final product, with peak scores of 4.41 and 4.66, respectively. The feature was investigated through Smart Formula Manually with the C 16 H 14 O 5 neutral molecular formula being assigned to this mass (mass accuracy of −3.2 ppm/−0.9 mDa and 0.4 ppm/0.1 mDa; isotope fit of 4.3 mSigma and 16.8 mSigma for the raw material and the final product, respectively). As 195 candidates were retrieved (using the neutral molecular formula and a mass accuracy threshold of 5 ppm) from COCONUT4MetFrag, the retention time prediction tool was used to prioritize the candidates. The dihydroxyflavanone isosakuranetin with predicted retention time of 7.9 min was the most plausible candidate. Six of the eight experimental MS/MS fragments were explained by MetFrag and the MS/MS similarity score for the raw material and the final product with the MassBank Record: BS003552 is 0.731 and 0.727, respectively. However, because the precise position of the hydroxyl moieties on the benzene rings cannot be determined, the highest level of identification that can be reached is 3. For a higher level of identification, the comparison to the corresponding reference standards is required.
One chromatographic peak was observed for m/z 359.3170/359.3170 at 14.7 for both raw material and final product, with peak scores of 10.7 and 9.58, respectively. The neutral molecular formula C 21 H 42 O 4 was assigned to this mass (mass accuracy of −4.9 ppm/−1.8 mDa and −4.9 ppm/−1.8 mDa; isotopic fit (27.5 mSigma and 28.7 mSigma for the raw material and the final product, respectively). Employing COCONUT4MetFrag, five candidate structures (neutral molecular formula and 5 ppm mass accuracy) were retrieved with the predicted retention time for monostearin being 13.3 min, and 12 of the 20 experimental fragments (raw material) and 9 of the 17 fragments (final product) were explained by MetFrag. The MS/MS spectra were then compared to the corresponding of the MoNA ID spectrum library: FiehnHILIC001606, and they were found to be in agreement (MS/MS similarity score of 0.732 for the raw material and 0.736 for the final product). As a result, monostearin reached confidence level 2a.
One peak was observed in the chromatograms of the samples (t R 5.5 min and 5.6 min for raw and product, respectively) for 137.0252/137.0258 m/z values. The neutral mass of C 7 H 6 O 3 was assigned to it (mass accuracy: −5.5 ppm/−0.7 mDa and −10.3 ppm/−1.4 mDa) with 19 plausible candidates (using a neutral molecular formula, a threshold accuracy of 5 ppm, and the COCONUT4MetFrag Database) through MetFrag. Since all the fragments of the possible candidates (two out of two) were interpreted, their prioritization criteria were the MetFrag score and the retention time prediction. Among these, four isomers scored the highest MetFrag score (1.00), i.e., [3,4-dihydroxybenzaldehyde (t Rpred = 3.7 min), 2,4dihydroxybenzaldehyde (t Rpred = 4.5 min), 2,5-dihydroxybenzaldehyde (t Rpred = 4.2 min), 2,3-dihydroxybenzaldehyde (t Rpred = 4.1 min). Since the exact position of these hydroxy groups on the benzene ring cannot be defined, the identification level remains at level 3, requiring more evidence.
One chromatographic peak eluting at 10.8 min for both raw material and final product was observed for the m/z values of 288.2904/288.2903, with peak scores of 5.45 and 9.11, respectively, assigned to the neutral molecular formula C 17 H 37 NO 2 with mass accuracy of −2.3 ppm/−0.7 mDa and −2.2 ppm/−0.6 mDa, and isotopic fit of 7.6/13.2 mSigma. Using MetFrag one candidate, i.e., C17-sphinganine was retrieved from COCONUT4MetFrag (using the neutral molecular formula and a mass accuracy criterion of 5 ppm). The interpretation of 15 of the 28 experimental fragments for the raw material and 18 of the 34 for the final product was achieved, and the predicted retention time (t Rpred = 11.1 min) was close to the experimental one. The MS/MS spectrum obtained, was compared to the corresponding entry available in the MoNA spectral library (MoNA ID: CCMSLIB00000579284) and found to be in agreement (MS/MS similarity score of 0.822 for the raw material and 0.751 for the final product) reaching the 2a level of identification confidence.
A peak (m/z value of 331.2850/331.2851) was observed eluting at 14.1 min for both the raw material and its final product, with peak scores of 3.51 and 3.05, respectively. The neutral molecular formula C 19 H 38 O 4 was assigned to it (−2 ppm/−0.7 mDa, −2.4 ppm/ −0.8 mDa, 18.4/23.2 mSigma) and searched using MetFrag in the COCONUT4MetFrag database, deriving seven candidate structures. Among the plausible candidates, 2-palmitoy lglycerol, glycerin palmitate, and 2,3-dihydroxypropyl 14-methylpentadecanoate had the highest MetFrag Score and the same number of interpretative MS/MS fragments (11 out of 22 fragments for onion and 11 out of 20 fragments for the final product). The predicted elution time was 12.7, 12.8, and 12.3 min respectively, with glycerin palmitate showing the smallest deviation from the experimental retention value between the candidates. This is reasonable, as the interaction of the glycerin palmitate chain is stronger than these with branches, delaying its elution time [32]. The experimental MS/MS spectrum of glycerin palmitate was compared to the corresponding one in the MoNA spectral library (ID CCMSLIB00000849055) and found to be in accordance (MS/MS similarity score of 0.698 for the raw material and 0.686 for the final product), reaching a confidence level of 2a.
One chromatographic peak was observed for the m/z 151.0401/151.0402 value at 4.8 min for the raw material and the final product, respectively. However, due to the low peak intensity, its molecular formula could not be matched. COCONUT4MetFrag was used to identify potential candidate structures with this m/z value and a mass accuracy threshold of 5 ppm. The m/z value corresponds to 67 candidates, with the proposed neutral molecular formula was C 8 H 8 O 3 . Among the candidate compounds was vanillin, whose predicted retention time (t Rpred = 4.7 min) and reference standard retention time (t R = 4.8 min) were in good agreement with the experimental one of the samples. Thus, this m/z value can be identified to a confidence level of 5.
For the theoretical value m/z 227.2017, a chromatographic peak eluting at 13.0 min (both raw and final product) was identified as a suspect compound (myristic acid) in our list. This m/z value (mass accuracy of 1.0 ppm/0.2 mDa, 0.4 ppm/0.1 mDa) isotopic fit (4.7 mSigma, 17.2 mSigma) was assigned to the neutral molecular formula C 14 H 28 O 2 for both samples. The compound's MS/MS spectrum matched that of MassBank ID: BS003878, reaching a 2a level of confidence. The standard's MS/MS spectrum and elution time (t R = 11.9 min) were compared to those of the samples and found to be in good agreement. Thus, its identification reached the highest level (level 1).
The feature corresponding to the m/z value 255.2330 (t R raw material = t R final = 13.8 min) is matched to the suspect compound palmitic acid. The neutral molecular formula C 16 H 32 O 2 was assigned to it, which met all the criteria (mass accuracy = 0 ppm/0 mDa, 1.2 ppm/ 0.3 mDa, isotopic fit = 13.1 mSigma, 3.2 mSigma) for both samples, and the predicted retention time (t Rpred = 12.9 min) was acceptably close to their corresponding experimental retention time. The candidate's MS/MS fragmentation matched that of the MoNA spectral library (MoNA ID: MetaboBASE1104) (level 2a). The samples' MS/MS spectra and retention time were compared to those of the standard (t R of standard = 13.8 min), and they were found to be in agreement. Therefore, palmitic acid reached confidence level 1.
For the m/z value 279.2330, a chromatographic peak was observed at 13.4 and 13.5 min for the raw material and the final product, respectively, corresponding to linoleic acid, included in the suspect list. The neutral molecular formula C 18 H 32 O 2 was assigned to this value (mass accuracy of 0.9 ppm/0.2 mDa, 2.4 ppm/0.7 mDa; isotopic fit of 13.1 mSigma, 22.7 mSigma) for the raw material and the final product, and the predicted elution time (t Rpred = 12.9 min) was acceptably close to the experimental ones. The MS/MS fragments were in line with those of the MassBank (MassBank ID: RP029511), reaching a level of confidence of 2a. The reference standard's MS/MS spectra and elution time (t R = 13.5 min) were identical to those of the samples. These steps confirm the identification of linoleic acid at level 1.
For the theoretical value m/z 281.2486, a chromatographic peak (t R raw = 13.9 min, t R final = 14.0 min) corresponding to the suspect compound oleic acid was noticed. The neutral molecular formula C 18 H 34 O 2 was assigned to this value (raw = 0.9 ppm/0.2 mDa, 35.1 mSigma, final = 1.7 ppm/0.5 mDa, 0.2 mSigma). Thus, it was confirmed at the 2a level of identification via the comparison of the experimental to the fragments from the reference standard's MS/MS spectrum in the mass spectral library (MassBank ID: BS003989). The standard's MS/MS spectrum and retention time (t R = 14.0 min) were found to be in good agreement with the samples' respective ones, resulting in the identification of oleic acid at level 1.
For the m/z value 283.2643, a chromatographic peak (t R raw = 14.4 min and t R final = 14.4 min), corresponding to the suspect compound stearic acid, was detected. The neutral molecular formula C 18 H 36 O 2 was assigned to this value, fulfilling the criteria of mass accuracy and isotopic fit (raw = 0.2 ppm/0.1 mDa, 3.6 mSigma and final = 1.5 ppm/0.4 mDa, 6.2 mSigma) for both samples. Comparing the fragments to those from the reference spectrum (MassBank ID: BS003930), as well as the agreement of the predicted (t Rpred = 13.4 min) to the experimental retention time, the 2a identification confidence level was reached. Additionally, stearic acid was confirmed by the reference standard and reached the highest level of identification (level 1).
A chromatographic peak was observed at 4.7 min (raw and final product) for the theoretical value m/z 121.0295, which corresponds to the suspect compound, 4-hydroxybenzaldehyde. The neutral molecular formula C 7 H 6 O 2 was assigned for the raw material and final product (mass accuracy: 3.8 ppm/0.5 mDa, 1.97 ppm/0.24 mDa, isotopic fit: 2.3 mSigma, 33.1 mSigma, respectively). The predicted elution time (t Rpred = 4.2 min) agrees to the experimental one, and its MS/MS was identified by mzCloud No ID: 234, reaching the 2a confidence level.
The theoretical value m/z 132.1019 for both samples eluting at 1.7 min belongs to a candidate (leucine) from the suspect list. This mass was assigned the neutral molecular formula C 16 H 13 NO 2 , which met the criteria of mass accuracy (3.3 ppm/0.4 mDa, 4.0 ppm/0.5 mDa) and isotopic fit (9.9 mSigma, 43.1 mSigma) for mushroom straw and its final product, respectively. The predicted retention time (t Rpred = 1.5 min) agreed with the experimental retention time of the samples. Leucine was confirmed at confidence level 2a by comparing its MS/MS spectrum with the one available in the spectral library (mzCloud No ID: 06). The MS/MS spectrum and the retention time of the standard (t R = 1.7 min) agreed with the corresponding ones of the samples, reaching level 1 A chromatographic peak eluting at 1.4 min (raw and final product) corresponds to the candidate compound choline from the suspect list with the theoretical m/z value of 104.1070. This value was assigned the neutral molecular formula C 5 H 14 NO (mass accuracy of −6.6 ppm/−0.7 mDa, −8.8 ppm/−0.9 mDa for the raw and final products). The predicted retention time (t Rpred = 1.7 min) matched that of the experimental retention time of the samples. Choline was verified by comparing its experimental MS/MS spectrum to the corresponding one in Mass Bank library (ID: PR100405), thereby, gaining the 2a level of confidence.
The theoretical value m/z 131.1291 eluting at 1.4 min for both samples, corresponds to the candidate compound agmatine. The neutral molecular formula C 5 H 14 N 4 was assigned to this value (mass accuracy of −6.5 ppm/−0.9 mDa, isotopic profile (4.9 mSigma) for the raw material. The predicted elution time (t Rpred = 0.9 min) was in accordance with the experimental one of the samples. Considering all the above, we concluded that the agmatine was identified by MoNA ID: PT106600 at a level of confidence of 2a.
A chromatographic peak was observed at 1.4 and 1.3 min for the raw material and final product, respectively, which corresponds to the suspect compound spermidine with a theoretical m/z value of 146.1652. This value was assigned the neutral molecular formula C 7 H 19 N 3 (mass accuracy of −3.4 ppm/−0.5 mDa for raw and final product, 8.9 ppm/1.3 mDa). The predicted retention time (t Rpred = 0.9 min) matches the samples' experimental retention time. Spermidine was verified by comparing its experimental MS/MS spectrum to the corresponding one in the MoNA spectral library (ID: MoNA032833), yielding a level of confidence of 2a.

Non-Target Screening of the Raw Straw Mushroom Material and Mushroom-Based Final Product
Two compounds (9-hydroxy-8,10-dehydrothymol, x,y-dihydroxy-benzaldehyde x = 3,2,2,2 y = 4,4,5,3) were identified through the non-target screening workflow, reaching at the confidence level 3. The notation followed for the mass spectral data corresponds to the experimental m/z value raw material /m/z value final product .
For the m/z values 137.0252 and 137.0258 for the onion raw material and mushroombased final product, respectively, one chromatographic peak eluted at 5.5 min. The identification process for this m/z value, which corresponds to x,y-dihydroxy-benzaldehyde (x = 3,2,2,2 y = 4,4,5,3), was discussed in Section 2.3, as this compound was derived to the mushroom-based final product as a result of the contribution of the onion raw material.

Beneficial Role of the Identified Compounds in Plant Health
Plants are constantly exposed to environmental stresses, which can be classified as biotic (e.g., pathogen attack by fungi, bacteria, and herbivores) or abiotic (e.g., drought, radiation, salinity, floods, and extreme temperature conditions) [46]. Two final products (onion-and mushroom-based) were developed to counteract the adverse effects of these conditions and promote plants health. The beneficial role of the specific compounds identified in the final products is briefly discussed.
Free fatty acids'(FFA) ability to kill or inhibit the growth of bacteria render them appealing as antibacterial agents, especially when the use of conventional antibiotics are not preferable or permitted [47]. Additionally, the inhibitory effect of long chain fatty acids against Gram-positive bacteria [48], as well as the involvement of plant-derived C18:0, C18:1, C18:2, and C18:3 (FFA) against a wide range of pathogens and opportunistic microorganisms have been discussed in the scientific literature [49]. Citric acid, belonging to the category of low molecular weight organic acids, is an effective agent for removing significant amounts of heavy metals (Cd, Cu, and Pb) from soil (known as metal phytoextraction) [50].
Moreover, flavonoids may act as attractants or feeding stimulators for some insect species as well as contribute to plant defense [51]. Specifically, quercetin promotes proper plant growth and development by facilitating a variety of physiological processes in plants, including seed germination, pollen growth, and photosynthesis. In addition to the aforementioned beneficial effects, quercetin may also aid in plant resistance to biotic and abiotic stresses due to its high antioxidant capacity [52]. Furthermore, Kurepa et al. recommended incorporating quercetin into growth media due to its beneficial effect on plant stress resistance [53].
Choline could facilitate in plant resistance to osmotic stress (drought and salinity), as well as nutritional value enhancement [54]. Spermidine, which acts as a plant growth regulator, contributes significantly to regulating plants' stress tolerance [55]. Thus, the presence of the above-mentioned compounds in the final products may improve plant health.

Compost Methodology
In the composting area of the IKORGANIC company production unit, a space was demarcated and cleaned to create composting piles (batch) comprised of chopped raw materials The procedure started with the creation of composting series and the diffusion of oxygen to them via perforated tubes. The temperature was monitored in a daily basis. An increase in temperature indicates the formation of mesophilic microorganisms, which are responsible for the initialization of material fermentation and the denaturation of the compounds thereof. Thermophilic microorganisms formed when the temperature rises above 45 • C.
After 6-10 days, when the homogenization process was completed, the pile was inverted and placed in a new batch. The heaps were frequently inverted to allow oxygenation, resulting in a constant rise in temperature and the completion of fermentation. The pile was inverted for the second time either at the end of 10-12 days or when the temperature reached 75 • C. The pile's temperature remained high for 5-10 days before gradually declining. Following the processing of the raw materials and the placement of the piles in the proper temperature-humidity-ventilation conditions, a temperature decrease was observed as cryophilic microorganisms grew.
When the temperature reached 40-50 • C, the material was removed from the composting square and placed on a waiting maturation area for 2 to 3 weeks, during which it was soaked in water and inverted, the composting process was completed, and the maturation stage for the samples began (Supplementary Figure S7). The samples were then transitioned to a curing area and covered with cloth for a period of time, allowing their composition to be stabilized. Several parameters, including the type (onion or mushroom) and homogeneity of the mixture, its humidity, how it is cut, and the external weather conditions, may influence the required time for compost competition. Following the maturation process, the products were passed through a sieve with a 2 cm diameter.

Onion-Based Final Product
The onion compost was hydrolyzed with addition of aqueous glycerin mixture (2:1, v/v) at 30 • C, yielding a liquid with biostimulant properties.

Mushroom-Based Final Product
The mushroom-based final product comprised of a mixture from the mushroom compost and the onion-based final product in a proportion of 4:1 (w/v). The abovementioned was homogenized thoroughly using a mixer. The homogenization procedure lasted 8 days at a constant temperature of 20 • C, and the resulting liquid was the mushroombased final product.
Copper (II) sulfate pentahydrate (CuSO4 * 5H 2 O 99%) and vanillin were purchased from Alfa Aesar (Karlsruche, Germany). Stock solutions of the reference standards (1000 mg L −1 ) were prepared in MeOH (LC-MS grade) and stored at −20 • C in ambient glass containers. Working solutions of various concentrations were prepared by appropriate dilutions of the stock solutions with a mixture of MeOH: H 2 O (50:50, v/v). A multi-element certified reference material (20 analytes including Zn, Cd, V, Fe, Ba, Cu, Ni, B, Pb, Mo, Mn, Cr, and Al) was bought by CPAchem (Zagora, Bulgaria).

Sample Pre-Treatment for HRMS Analysis
The onion and mushroom samples were freeze-dried to achieve compound preconcentration and better preservation. Various parameters affecting the yield of the solid-liquid extraction used as the pretreatment methodology were tested, i.e., the solvent polarity (MeOH, acetone, hexane, and H 2 O in various proportions), the use of extraction assisted sonication, the extraction time and temperature. The description of a sample preparation procedure of the raw materials and their corresponding products for the HRMS analysis can be found in detail in the main text (Sections 4.4.2 and 4.4.3), while the other sample preparation procedures are detailed in the Sections A.2.6.1 to A.2.6.4.

Freeze-Drying of Raw Materials
The onion samples were homogenized after being chopped into little pieces using a home homogenizer, whilst the mushroom straw was cut into smaller pieces with a pair of scissors. Following, the samples (onion with peel, mushroom straw) were weighed in pre-weighed petri dishes and stored at −80 • C for five hours in order to facilitate freezedrying. The temperature of the freeze drier was set to −55 • C, and the obtained vacuum was 0.05 mbar. The procedure was completed after 24 h, and the lyophilized samples in the petri dishes were weighed again. The percentage of moisture for the onion and mushroom straw were (90.9 ± 0.1)% and (90.0 ± 0.3)%, respectively.

Sample Preparation of Onion and Its Corresponding Final Product
The same sample preparation was followed for both raw material and its corresponding final product, differentiating only the initial mass of the investigated sample. Therefore, either 0.5 g of a lyophilized onion sample or 2.0 g of the onion-based final product were weighed in 50-mL centrifuge tube. The addition of 10 mL (MeOH: aq 1% formic acid, 80:20 (v/v)) was followed. Subsequently, they were vortexed for 1 min, followed by shaking employing a horizontal shaker for 30 min. Afterwards, the samples were placed in an ultrasonic bath at 30 • C for one hour and centrifuged at 4000 rpm for ten minutes. The supernatants were then collected, filtered through RC syringe filters, and transferred to 2-mL autosampler glass vials. The extracts were injected into the UPLC-QToF-MS system.

Sample Preparation of Mushroom Raw Material and Its Corresponding Final Product
Considering the higher concentration of bioactive compounds in onions than in mushroom samples reported in the literature [25,28,39,56], a preconcentration step was added to the above-mentioned procedure. As a result, the aforementioned extraction step was performed for these samples, followed by an evaporation step. The supernatants were collected in round-bottom flasks and evaporated at 35 • C using a rotor evaporator. The residues were reconstituted with the addition of 5 mL of EtOH, transferred to glass tubes and evaporated to dryness under a gentle nitrogen stream at 35 • C. Finally, the residues were reconstituted with the addition of 0.5 mL of MeOH: H 2 O (50:50, v/v), and the extracts were filtered through RC syringe filters, transferred to 2-mL autosampler glass vials, and injected into UPLC-QToF-MS system.
The column temperature was kept constant throughout the chromatographic analysis at 30 • C, and the injection volume was set at 5 µL. The analysis was conducted in both positive and negative ionization modes. The mobile phases in the positive ionization mode were composed of: (A) aq. 5 mm ammonium formate: MeOH (90:10 v/v) acidified with 0.01% formic acid and (B) 5 mM ammonium formate in MeOH acidified with 0.01% formic acid and, in the negative ionization mode, were (a) aq. 10 mM ammonium acetate: MeOH (90:10 v/v) and (B) 10 mM ammonium acetate in MeOH. The same gradient elution program was used in both ionization modes.
Initially, the column was equilibrated at 99% (A) at a flow rate of 0.2 mL min −1 for 30 min. The elution program was started with 99% (A), with a flow rate 0.2 mL min −1 and remained constant for 1 min. Afterward, it was decreased to 61% in 2 min without changing the flow rate. The flow rate was increased to 0.4 mL min −1 over the next 11 min, while (A) was reduced to 0.1%. In the next step, the flow rate was increased to 0.48 mL min −1 , while (A) was kept constant in the ratio of 0.1%.
Finally, the ratio of the aqueous (A) to organic (B) phases as well as the flow rate were restored to the initial conditions for 3 min before the next injection. Regarding the MS parameters, the values 3500 V for the capillary voltage, 2 bar for the nebulizer gas pressure (N 2 ), 8 L min −1 for the drying gas, and 200 • C for the capillary temperature were selected.
The Q-ToF system was calibrated on a daily basis using sodium formate as the calibrant, which was prepared in H 2 O: isopropanol (50:50, v/v) and injected at the beginning of each run. Both data independent (broad-band Collision Induced Dissociation -bbCID: DIA) and data dependent acquisition (AutoMS, DDA) modes were used for the analysis employing two MS methods. The MS and MS/MS spectra were acquired by applying two different collision energies (4 eV and 25 eV) in the bbCID mode. The five most abundant ions per MS scan were chosen and fragmented with ramp collision energy (collision energy based on m/z values) in AutoMS mode.

Instrumentation for Additional Analyses
The instrumentation used for the physicochemical (conductivity, pH), the nitrogen and the carbon determination, as well as the elemental analysis, can be found in the Appendix A.3.

Identification Confidence
The feature annotation was performed according to the Schymanski et al. scheme, considering the five levels of confidence in identifying a plausible candidate [57]. Starting at confidence level 5, the exact mass of interest was confirmed, although no other evidence for the candidate's molecular formula exists. The feature's formula was assigned at the next level of identification confidence (level 4) [58]. At the next level of identification confidence (level 3), a plausible candidate was tentatively identified by evaluating its MS/MS fragments using in silico fragmentation tools (MetFrag [59] or CFM-ID [60]) and its identification was enhanced using prioritization methods, such as retention time prediction [32] and ionization efficiency estimation [61].
The existence of a diagnostic ion in a potential candidate may result in identification at level 2b, whereas level 2a is reached when the similarity score between the MS/MS candidate's spectra and its corresponding spectrum available online spectral libraries is greater than 0.7. Finally, the highest level of identification confidence (level 1) was achieved when the reference spectra and retention time of the candidates were in accordance with those of the reference standard.

Suspect Screening Methodology
The approach of suspect screening is based on 'prior knowledge' of the compounds present in the sample considering the corresponding scientific literature. A suspect list was compiled, comprising compounds identified in onions according to the literature using various analytical techniques [13][14][15][16][17][18][19]21,[33][34][35][36] A quantitative structure-retention (QSRR) model was also used to predict the plausible candidate's retention time [32,62]. The initial level for a compound identification in the suspect screening approach is level 3 (level of identification confidence 3), based on the relative confidence scheme reported in the literature [57]. The mass accuracy (expressed in ppm or mDa) and the isotopic profile accuracy (expressed in mSigma) [63,64] were evaluated using the 'Smart Formula' software (which is part of Data Analysis, Bruker Daltonics, Bremen, Germany).
The molecular formula was assigned to the mass value (m/z) by mass tolerance of 2 mDa and isotopic tolerance of 100 mSigma. The MetFrag in silico fragmentation software [65] was used to assist the identification of the compounds by interpreting MS/MS fragments of this compound [65]. The MS/MS spectra of suspect compounds were searched in the literature and spectral libraries (MoNA [66], MassBank [67], GNPS [68], and mz-Cloud [69] (confidence level 2). A QSRR, model was also used to predict the candidate molecule's retention time [32,62] developed in-house based on the proposed structure, physicochemical properties, and chromatographic system (analytical column, elution program and mobile phases). The elution time was predicted for all compounds that reached the 2a, 2b, and 3 confidence levels, and the compounds were classified into a different box (an example is illustrated in Figure 1b) based on the criteria listed in the literature [32]. Finally, the MS/MS spectra of the suspect compounds were compared to those of the authentic reference standards (confidence level 1), if available in the laboratory. Since the analytical data available to identify the targeted m/z values varied, the identification reached different levels of confidence.

Non-Target Screening Methodology Peak Picking and Initial Screening
Non-targeted screening was performed to identify common compounds found in raw materials and their corresponding final products using the open-source MS Dial software (version 4.60) [70]. Initially, the raw data was calibrated in Data Analysis software (Bruker Daltonics, Bremen, Germany) and converted by the ABF converter [71]

Peak Annotation
Peak annotation was performed to assign the detected features to plausible candidates. The workflow is described below. Peak scores, defined as the ratio of the candidate peaks area to its corresponding peak intensity, have been calculated for the m/z values of interest. A value > 4 is considered acceptable [63]. Following that, the assigned formula's mass accuracy and isotopic profile were evaluated using the Smart Formula Manually software (Bruker Daltonics, Bremen, Germany) to achieve a level of confidence 4. The candidates (using the molecular formula of the neutral species showing ≥ 5 ppm mass accuracy) were retrieved from the COCONUT4MetFrag database and proceeded to MetFrag for obtaining the in silico MS/MS predicted spectra.
It is worth noting that COCONUT (the Collection of Open Natural Products) is an open database of 406.744 natural products [73]. The candidates were then prioritized using an in-house retention time prediction tool (confidence level 3) [32,63]. The MS/MS spectra of the suspect compounds were then searched against the literature and mass spectra libraries (MoNA [66], MassBank [67], GNPS [68], and mzCloud [69] confidence level 2). In cases where neither the MS/MS spectrum could be retrieved from the online libraries nor the standard compound was available, in silico fragmentation was performed, and the most abundant explained fragments were evaluated.
The MS/MS similarity score for each compound should be greater than 0.7 in order to be annotated. The elution time was predicted for all compounds that reached the 2a, 2b, and 3 confidence levels, and the compounds were classified into a different box (an example is illustrated in Figure 1b) based on the criteria listed in the literature [32]. Finally, where reference standards were available in the laboratory, MS/MS spectra were compared to those of suspect compounds (confidence level 1), with similarity score > 0.7.

Conclusions
The composting process was applied for the production of two biofertilizers using onions and mushroom by-products as raw materials. These two raw materials and their corresponding final products (onion-based and mushroom-based) were characterized following suspect and non-target methodologies through UPLC-QToF-MS, and the common compounds were elucidated between the investigated samples. In total, 14 compounds were identified in both the onion raw material and the onion-based final product, while 12 compounds were found in both mushroom and the mushroom-based final product. In the latest mentioned product, the contribution of the onion was also taken into consideration as it was a product component.
In both cases, the identified compounds, belonging to various categories, such as fatty acids, organic acids, flavonoids, and amino acids, enhancing their beneficial effects to plant health. The fatty acids appeared to be the most resilient compounds during the compost process as the majority of the identified compounds belong to this class, while antioxidants appeared to decrease as expected. Additionally, parameters, such as the pH, conductivity, organic matter, nitrogen content, and elemental analysis, were determined for the overall characterization of the aforementioned final products. The results indicate that the presence of the identified compounds could ameliorate plant health.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/molecules27113498/s1, Figure S1: Typical chromatograms of myristic acid (227.2017 ± 0.005 Da), linoleic acid (279.2330 ± 0.005 Da), palmitic acid (255.2330 ± 0.005 Da) and oleic acid (281.2486 ± 0.005 Da) for (a) the onion raw material and (b) the onion-based final product, Figure S2: Typical chromatograms of quercetin (301.0354 ± 0.005 Da) for (a) onion raw material and (b) the onion-based final product, Figure S3: Typical chromatograms of (a) isosakuranetin (285.0768 ± 0.005 Da) (b) x-y dihydroxybenzaldehyde (137.0244 ± 0.005 Da) (c) fumaric acid (115.0037 ± 0.005 Da) for the raw material and its corresponding final product, Figure S4: Typical chromatograms of the myristic acid (227.2017 ± 0.005 Da), linoleic acid (279.2330 ± 0.005 Da), palmitic acid (255.2330 ± 0.005 Da), oleic acid (281.2486 ± 0.005 Da) and stearic acid (283.2643 ± 0.005 Da) for (a) the mushroom raw material and (b) mushroom-based final product, Figure S5: Typical chromatograms of the agmatine (131.1291 ± 0.005 Da) for (a) the mushroom raw material and (b) the mushroom-based final product, Figure S6: Typical chromatograms of choline (104.1070 ± 0.005 Da) for the mushroom raw material and the mushroom-based final product, Figure S7. Temperature versus time of composting (a) for the onion based final product (b) for the mushroom-based final product. Table S1. Results of analysis in onion-based and mushroom-based final product.  to 50 mL volumetric flasks and diluted to volume with ultrapure water. A five-point calibration curve, ranging from 0.1 mg L −1 to 5 mg L −1 , was constructed for quantification purposes by dilutions of the stock solution with ultrapure water.
Appendix A.2.5. Phosphorus Determination A photometric determination was performed for the determination of phosphorus in accordance with Regulation (EC) No 152/2009. The sample was digested as described in Appendix A.2.4. For quantification purposes, the construction of a five-point calibration curve ranging from 0.5 to 10 mg L −1 was performed, while appropriate dilutions were conducted. The wavelength was set at 430 nm.
Appendix A.2.6. Sample Preparation for HRMS Analysis Different sample preparations were performed to achieved the highest number of features and the highest extraction efficiency of the analytes expressed as the most intense signal. All the described procedures have been applied to the samples and all the data resulting for these procedures were evaluated by the suspect and non-target workflows. Two additional sample preparations for the raw materials and the final products are presented in Appendices A.2.6.1-A.2.6.4.

Raw Material's Sample Preparation for HRMS Analysis
In 50-mL centrifuge tubes, 0.5 g of lyophilized samples (onions, mushroom straw) were weighed. Following that, 10 mL (MeOH: H 2 O, acidified with 1% formic acid) were added in a proportion of 50:50 (v/v). After stirring for 1 min in a vortex device, the samples were placed in a 65 • C thermostated waterbath for 4 h. The samples were shaken for 16 h on a horizontal shaker before being centrifuged for 10 min at 4000 rpm. Afterwards, the supernatants were collected and filtered through RC filters and transferred to 2-mL autosampler glass vials. The extracts were injected into the UPLC-QToF-MS system.

Final Products' Sample Preparation for HRMS Analysis
In 50-mL centrifuge tubes, 2.5 g of final products were weighed and the addition of 2.5 mL of MeOH acidified with 1% formic acid followed. The samples were stirred in a vortex device for 1 min, placed in a water bath at 65 • C for 4 h and then shaken for 16 h on a horizontal shaker. The samples were centrifuged at 4000 rpm for 10 min, and the supernatants were filtered through RC filters and transferred to 2-mL autosampler glass vials. The extracts were injected into the UPLC-QToF-MS system.

Raw Materials' Sample Preparation for HRMS Analysis
In 50-mL centrifuge tubes, 0.5 g of the lyophilized samples was weighed. After moisturizing the matrices with 5 mL of acidified H 2 O (1% acetic acid), 5 mL of hexane were added. Afterwards, the samples were vigorously shaken for 30 min on a horizontal shaker, mixed for 1 min in a vortex device, and placed in an ultrasonic bath for 30 min at 30 • C. Subsequently, the samples were centrifuged at 4000 rpm for 10 min and the supernatants were collected in glass test tubes. The above-mentioned procedure was repeated once. The collected supernatants (organic layer) were evaporated under a gentle nitrogen stream at 35 • C, and the residues were reconstituted with 250 µL (MeOH: H 2 O) at a proportion of 50:50. The reconstituted extracts were filtered using RC syringe filters and transferred to 2-mL autosampler glass vials. The extracts were injected into the UPLC-QToF-MS system.

Final Products' Sample Preparation for HRMS Analysis
In a 50-mL centrifuge tube, 2.5 g of each final product were weighed and the addition of 2.5 mL of hexane was followed. The samples were stirred vigorously in a vortex device for 1 min and then shaken on a horizontal shaker for 30 min. The samples were then placed in an ultrasonic bath for 30 min at 30 • C. The samples were centrifuged at 4000 rpm for 10 min and the organic layer was collected in new glass test tubes. The abovementioned procedure was repeated once. The evaporation of the collected supernatants was carried out under a gentle nitrogen stream at 35 • C. Afterwards, reconstitution of the residues with the addition of 250 µL (MeOH: H 2 O) at a ratio of 50:50 (v/v) was performed. The reconstituted extracts were filtered using RC syringe filters and transferred to 2-mL autosampler glass vials. The extracts were injected into the UPLC-QToF-MS system.