Applying LFQRatio Normalization in Quantitative Proteomic Analysis of Microbial Co-culture Systems

Materials and reagents

Biological materials

1. Bacterial monoculture 1: Synechococcus elongatus PCC 7942 cscB/SPS monoculture (from Abramson et al. [12]; kindly provided by Prof. Daniel Ducat). For growth monitoring information, see General notes

2. Bacterial monoculture 2: Azotobacter vinelandii DJ ΔnifL monoculture (from Ortiz-Marquez et al. [13]; kindly provided by Prof. Leonardo Curatti)

3. Bacterial co-culture (combination of bacterial monocultures 1 and 2): Synechococcus elongatus cscB/SPS and Azotobacter vinelandii ΔnifL co-culture

Caution: These strains are genetically modified and so need to be handled and disposed of according to the recommended safety guidelines of your institute.

Reagents

1. Sodium dodecyl sulfate (SDS) (Sigma-Aldrich, catalog number: 71725)

2. Phosphate-buffered saline (PBS) tablets (Millipore, catalog number: 6500-OP)

3. 100× Halt protease inhibitor cocktail (Thermo Scientific, catalog number: 78446)

4. 425–600 µm glass beads, acid-washed (Sigma-Aldrich, catalog number: G8772)

5. Urea, proteomics grade (Thermo Scientific, catalog number: 29700)

6. Trizma^® base (Sigma-Aldrich, catalog number: T1503)

7. Buffers for adjusting pH (e.g., NaOH/HCl) (Sigma-Aldrich, catalog number: 06203/258148)

8. DL-Dithiothreitol (DTT) (Promega, catalog number: V3151)

9. Water, HPLC grade (Fisher, catalog number: W/0106/PB17)

10. Iodoacetamide (IAA) (Sigma-Aldrich, catalog number: 1149)

11. Calcium chloride (CaCl₂) (Sigma-Aldrich, catalog number: C3306)

12. Trifluoroacetic acid, HPLC grade (TFA) (Sigma-Aldrich, catalog number: 15608360 or similar)

13. Sequencing-grade modified trypsin (Promega, catalog number: V5111)

14. Formic acid, LC-MS grade (Thermo Scientific, catalog number: 85178)

15. Acetonitrile (ACN), HPLC grade (Fisher, catalog number: A/0627/17)

16. Methanol (MeOH), HPLC grade (Fisher, catalog number: M/4056/17)

Solutions

1. 1 M Tris-HCl (pH 8.5) (see Recipes)

2. 1 M DTT (see Recipes)

3. Lysis buffer (see Recipes)

4. Urea buffer (see Recipes)

5. 100 mM iodoacetamide (IAA) (see Recipes)

6. 1 mg/mL trypsin stock (see Recipes)

7. 100 mM CaCl₂ (see Recipes)

8. 50 mM Tris-HCl (pH 8.5)/10 mM CaCl₂ (see Recipes)

9. MS loading buffer (see Recipes)

10. 80% ACN (see Recipes)

11. MS mobile phase (see Recipes)

Recipes

Note: Storage conditions are room temperature unless otherwise stated.

1. 1 M Tris-HCl (pH 8.5)

Weigh 12.114 g of Tris and dissolve in 80 mL of HPLC-grade water. Lower the pH to 8.5 with HCl and top up to 100 mL with HPLC-grade water.

2. 1 M DTT

Dissolve 1.542 g of DTT in 10 mL of HPLC-grade water. Aliquot and store at -20 °C or use immediately.

Note: DTT oxidizes in air, so the solution must be prepared fresh or frozen aliquots must be used immediately and then discarded.

3. Lysis buffer

4. Urea buffer

Note: We recommend using 1.5 mL LoBind Eppendorf tubes to prepare the solution. Always make up fresh urea solution. When in solution, urea degrades within a few hours into a compound that can alter the composition of your protein sample.

5. 100 mM iodoacetamide (IAA)

Dissolve 0.0185 g of iodoacetamide in 1 mL of HPLC-grade water.

Note: Always prepare fresh as IAA is light-sensitive. Wrap in foil once prepared.

6. 1 mg/mL trypsin stock

Add 20 µL of resuspension buffer (provided by the trypsin manufacturer) to one vial of lyophilized sequencing-grade modified trypsin (20 µg) to generate the trypsin stock. Store at -80 °C for up to 1 month.

7. 100 mM CaCl₂

Dissolve 1.47 g of CaCl₂ in 100 mL of HPLC-grade water.

8. 50 mM Tris-HCl (pH 8.5)/10 mM CaCl₂

9. MS loading buffer

Caution: Prepare in a fume cabinet.

10. 80% ACN

Caution: Prepare in a fume cabinet.

11. MS mobile phase

Solvent A

Solvent B

Caution: Prepare in a fume cabinet.

Laboratory supplies

1. 1,000 µL pipette tips

2. 200 µL pipette tips

3. 10 µL pipette tips

4. 50 mL sterile conical tubes (Falcon, catalog number: 352070)

5. 1.5 mL protein LoBind tubes (Eppendorf, catalog number: 0030108116)

6. 2D Clean-Up kit (Cytiva, formerly GE Healthcare Life Sciences, catalog number: 80-6484-51)

7. Pierce^TM C18 spin columns (Thermo Scientific, catalog number: 89870)

8. 0.3 µL PP snap ring HPLC vials (VWR, catalog number: 548-0120A)

9. ND11 open-top PE HPLC vial snap caps (VWR, catalog number: 548-3203A)

Equipment

1. Pipettes (p1000, p200, p10, p2.5) (Eppendorf)

2. Spectrophotometer (Jenway, model: 6305)

3. Temperature-controlled centrifuge (e.g., Thermo Scientific, model: SL 1ER)

4. Vortex with attachment for multiple microfuge tubes (Scientific Industries, model: Vortex-Genie 2)

5. Heated water bath (e.g., Grant, model: JB Nova)

6. Ultrasonic water bath (Fisherbrand, model: FB15051)

7. Nanodrop spectrophotometer (Thermo Scientific, model: Nanodrop 2000)

8. Incubator capable of maintaining 30–37 °C (e.g., Thermo Scientific, model: MaxQ 6000)

9. Vacuum concentrator (Eppendorf, model: Concentrator Plus)

10. Nano-flow liquid chromatography (Thermo Fisher Scientific, model: U3000 RSLCnano)

11. Hybrid quadrupole-orbitrap mass spectrometer (Thermo Fisher Scientific, model: Q Exactive HF)

12. Easy-Spray C18 column, 75 µm × 50 cm, Pepmap RSLC, 2 µm, 100 Å (Thermo Scientific, catalog number: E5903)

Software and datasets

1. Uniprot (https://www.uniprot.org/, 02/03/2020)

2. MaxQuant (2.0.3.0)

3. RStudio (4.2.0)

4. LFQ-Analyst (https://analyst-suite.monash-proteomics.cloud.edu.au/apps/lfq-analyst/, 08/09/2023)

5. All code has been deposited to GitHub: https://github.com/Mengxun-Shi/bio-protocol (01/13/2025)

Procedure

A. Protein extraction

1. Grow cell cultures to a growth phase appropriate for your experiment and measure the optical density at the wavelength suitable for your culture. Centrifuge 20 mL of cells at 4,000× g for 10 min at 4 °C in 50 mL centrifuge tubes.

Note: We grew our co-cultures in 500 mL flasks over an 8-day period, sampling on Days 0, 4, and 8 [11] (see General note 1). However, we encourage the reader to select their harvest points based on their own biological questions and comparisons.

2. Wash the pellets by resuspending in 10 mL of PBS buffer and repeating the centrifugation. Pellets can be stored at -20 °C for up to 1 month before processing.

3. Resuspend the pellet in 500 µL of lysis buffer per 1 OD unit (e.g., if OD = 2, add 1 mL of lysis buffer) and add the appropriate amount of 100× Halt protease inhibitor cocktail (e.g., if the total volume of the pellet + lysis buffer is ~700 µL, add 7 µL of protease inhibitor).

4. Freeze the cells at -80 °C for >16 h and quickly thaw them in a water bath at 37 °C for 3–5 min to allow partial cell breakage. Immediately transfer to ice after thawing.

Note: The frozen pellets can remain at -80 °C in lysis buffer for ~1 month.

5. Label and pre-cool 1.5 mL LoBind microcentrifuge tubes on ice. Transfer the lysed cell sample to the LoBind tubes. For the remainder of the procedure, keep the samples on ice.

6. Add approximately 500 µL of 425–600 µm acid-washed glass beads to the samples, leaving 2–3 mm of cell suspension above the level of settled beads.

7. Break the cells by vigorous vortex mixing of the tube 20 times in cycles of vortexing for 30 s and cooling on ice for 30 s using a vortex with a multi-tube attachment.

8. Incubate samples at 95 °C for 5 min using a dry heat block.

9. Collect the supernatant in a fresh 1.5 mL LoBind tube following centrifugation at 15,000× g for 10 min at 4 °C. The crude protein sample can be stored at -20 °C for ~1 month.

Note: At this point, it would be a good idea to quantify your crude proteins. Our crude protein concentration was typically 2–3 mg/mL. This can be done using, for example, a detergent-compatible Bradford reagent (Pierce, catalog number: 23246). It is also advisable to run the protein samples on an SDS-PAGE gel.

10. Purify the crude samples using a 2D Clean-Up kit following the manufacturer’s instructions to remove interfering substances and, if required, to concentrate the proteins. The resuspension buffer will be the urea buffer (see Section B).

Caution: Note the hazard statements in the 2D Clean-Up kit for the precipitant and wash buffer, which include “causes severe burns and eye damage” and “suspected to cause cancer.”

Note: At the wash step, the protein pellet can be stored in wash buffer and wash reagent for up to 1 week at -20 °C. However, once the pellet has been dried for <5 min, it is advisable to immediately proceed to the digestion steps and resuspend the pellet in urea buffer.

B. Protein digestion

1. Dissolve protein pellet in 50 µL of urea buffer. Sonicate the sample in an ultrasonic water bath to solubilize the proteins; the suspension should be transparent.

Note: If the sonication step takes longer than 10 min, cool the samples on ice for 5 min and continue the sonication until the samples are transparent.

2. Take 1 µL of resuspended protein pellet and dilute 1:10 by adding 9 µL of H₂O. Estimate protein concentration using the Nanodrop: Use 1 µL of 1:10-diluted urea buffer as a blank. Add 1–2 µL diluted protein sample to the Nanodrop to estimate protein concentration at 280 nm.

3. Reduction step: Transfer 50 µg of protein to a fresh 1.5 mL LoBind tube. Dilute to 10 µL with urea buffer. Incubate at 37 °C in a water bath for 30 min.

Note: DTT, the reducing agent, is included in the urea buffer.

4. Alkylation step: Add 1.5 µL of 100 mM IAA to the protein solution and incubate in the dark at room temperature for 30 min.

5. Digestion step: Add 1 µL of 1 mg/mL sequencing-grade modified trypsin stock to the protein sample and then add 67.5 µL of 50 mM Tris-HCl (pH 8.5)/10 mM CaCl₂ to a total volume of 80 µL. Digest overnight at 37 °C.

Note: The protease:protein ratio recommended by the trypsin manufacturer is 1:100 to 1:20 (w/w); here, we used 1:50.

6. Terminate the digestion: Add 0.8 µL of formic acid to a final concentration of 1% (recommend 0.1–1% v/v). Once added, it can be stored at -20 °C for up to 48 h.

7. Use C18 Spin columns to purify the digested peptides and remove contaminants following the manufacturer’s instructions. Purified samples should be dried using a vacuum concentrator and stored at -20 °C (days/weeks) or -80 °C (months) until mass spectrometry analysis.

Note: The binding capacity of the C18 Spin columns is 30 µg. By overloading the column with 50 µg of peptides, 30 µg will reliably be captured and eluted without peptide loss.

C. Mass spectrometry

1. Dissolve dried peptide pellets in 30 µL of loading buffer for a peptide concentration of 1 µg/µL. Sonicate samples in a water bath for 3 min at room temperature to suspend the samples.

2. Centrifuge the sample at 15,000× g for 2 min. Take 5 µL from the top of the centrifuged sample and transfer to an HPLC vial. Top up with 15 µL of loading buffer (20 µL total volume; 250 ng/µL peptide concentration), ensuring not to introduce bubbles to the vial.

3. Inject 500 ng of purified peptides (2 µL) into the LC–MS system with Easy-Spray C18 column. Perform LC–MS/MS by nanoflow liquid chromatography coupled to a hybrid quadrupole-orbitrap mass spectrometer or equivalent LC–MS/MS equipment. Here, peptides were separated by reverse-phase high-performance liquid chromatography using an Easy-Spray C18 column (75 µm × 50 cm) at a flow rate of 300 nL/min and column temperature of 40 °C. Two mobile phases were used: mobile phase A was composed of water with 0.1% formic acid v/v, and mobile phase B was composed of 80% acetonitrile in water, 0.1% v/v formic acid. Both were mixed over time for gradient elution. A 2-step gradient was employed: from 3% B to 10% B over 5 min, then from 10% B to 50% B (0.1% formic acid in 80% acetonitrile) to 75 min.

4. The mass spectrometer was programmed for data-dependent acquisition with 10 product ion scans (resolution 30,000, automatic gain control 1 × 10⁵, maximum injection time 60 ms, isolation window 2.0 m/z, fixed first mass 100 m/z, normalized collision energy 28, and intensity threshold 3.3 × 10⁴) per full MS scan (resolution 120,000, automatic gain control 1 × 10⁶, maximum injection time 60 ms, scan range 375–1,500 m/z) with a 35 s dynamic exclusion time. Make sure to include both biological (≥3) and technical (>2) replicates and ensure the appropriate control samples are included for meaningful sample comparison(s) to ensure the biological question is addressed.

Data analysis

A. Protein identification and quantification

1. Create a reference database (FASTA format) using all protein sequences of both strains; in this case, S. elongatus PCC 7942 and A. vinelandii DJ appended with the amino acids sequences of sucrose permease (CscB) from Escherichia coli and sucrose phosphate synthase (SPS) from Synechocystis sp. PCC6803 from Uniprot (Dataset S3).

Note: This can be done with a simple notepad program, saving the file name with the appendage .fasta.

2. Load the raw MS data files and database to MaxQuant (2.0.3.0) [7].

3. Select modifications: Oxidation (M) and Acetyl (Protein N-term) for variable modification, and Carbamidomethyl (C) for fixed modification.

4. Select Trypsin/P for enzymatic digestion.

5. Filter peptide-spectrum matches and protein identifications using a target-decoy approach at a false discovery rate (FDR) of 1%.

6. Select LFQ and intensity-based absolute quantification (iBAQ) options for quantitative analysis. Other parameters are shown in the supplementary files (Table S1).

7. If your synthetic co-culture is a novel combination of two bacteria, and this is the first proteomics experiment conducted for this consortium, continue to Section B. Otherwise, proceed to Section C.

B. Assessment of factors relating to protein identification for first-time proteomic analyses of new synthetic co-cultures

The physicochemical characteristics of proteins, such as the range of pI [14], MW [15], and hydrophobicity [16], as well as the dynamic range distribution of protein abundances within the proteome [17], can affect factors such as protein extraction efficiency and ionization efficiency in the mass spec, and hence how many peptides are detected. Similarly, a much larger proteome or highly abundant protein could mean a bias toward identifying peptides from a specific organism [18]. Finally, shared peptides between the strains will conflict with quantitative data and so should be taken into account [19]. We therefore advise researchers to identify variations in these factors between co-culture species, which could affect quantitative accuracy and therefore the biological interpretation of LFQ proteomics data.

1. Preliminary physicochemical characteristics analysis

a. Assess the theoretical pI ranges of the proteins in S. elongatus and A. vinelandii using R script (File S1).

b. Assess the MW of the proteins in S. elongatus and A. vinelandii using R script (File S2).

c. Assess the hydrophobicity of proteins in the S. elongatus and A. vinelandii databases via the grand average of hydropathy (GRAVY) scores using R script (File S3).

d. Dynamic range: Quantify the abundance of detected S. elongatus and A. vinelandii proteins by absolute quantification (iBAQ). Use the values from MaxQuant outputs–ProteinGroups.txt file, calculating log10 iBAQ intensity to show the dynamic range of both strains. Plot log10 iBAQ intensity against protein number.

Note: The dynamic range is the range of MS1 peak intensities over which peptides can be detected. A wide dynamic range or a protein that is extremely abundant in one strain may affect the quantification of the co-culture proteome. Therefore, it is critical to examine the dynamic range of proteins for each strain in the microbial community before further analysis [11].

2. Proteome size analysis

a. Run MaxQuant multiple times against S. elongatus and A. vinelandii individual and merged databases. For example, we run MaxQuant four times: i) S. elongatus monoculture against the S. elongatus database (Dataset S1); ii) S. elongatus monoculture against the merged database (Dataset S3); iii) A. vinelandii monoculture against the A. vinelandii database (Dataset S2); and iv) A. vinelandii monoculture against the merged database (Dataset S3).

b. Compare the identified protein numbers of S. elongatus and A. vinelandii taken from proteinGroups.txt files of each run. Filter for proteins with two or more unique peptides.

Note: We expect the same strain to have the same/similar number of proteins identified in both individual and combined database runs, which means that database size does not affect protein identification in this case. However, this should be considered in co-culture proteomics workflows, particularly where proteome sizes of co-culture members vary more widely.

3. Shared peptides analysis

a. Calculate theoretical shared peptides using R script (File S4) of tryptic peptides of the specified size (8–25 amino acids).

b. Analyze measured shared peptides by comparison of the resulting peptide sequences of each strain from MaxQuant outputs–Peptides.txt file.

C. LFQRatio normalization and proteomic analysis

1. Open the ProteinGroup.txt file output from MaxQuant and select identified proteins with two or more peptides.

2. Normalize LFQ intensity data using the LFQRatio normalization method as shown in the following equation, i.e., for each individual protein detected, its LFQ intensity is divided by the sum of all protein LFQ intensities for its respective strain:

N o r m a l i z e d L F Q i n t e n s i t y = L F Q i n t e n s i t y o f o n e p r o t e i n t o t a l L F Q i n t e n s i t i e s o f t h e s t r a i n

3. Use the normalized proteomics data to analyze differentially expressed proteins using an LFQ proteomics analysis platform such as LFQ-Analyst [20]. The following steps will be for using LFQ-Analyst; however, other platforms can be used (see General notes).

4. Load two input files: the LFQRatio-normalized proteinGroups.txt file and an experimental design file (an exemplary experimental design file can be downloaded from the LFQ-Analyst website).

5. In the Advanced Options, the adjusted p-value cutoff and Log2 fold change cutoff can be selected; we applied the default options of 0.05 and 1, respectively.

6. The default imputation type (Perseus type) can be applied; however, this is not necessarily recommended (see General notes and troubleshooting)

7. The type of FDR correction can be selected; we applied Benjamini–Hochberg for our experiment.

8. A list of differentially expressed proteins for each pair-wise comparison will be generated, as well as quality control plots, volcano plots, and expression heatmaps. Example volcano plots showing differential expression data generated with and without application of LFQRatio are shown in Figure 2.

9. Other parameters are shown in Table 1.

Table 1. Parameter settings of LFQ-Analyst used for differentially expressed proteins analysis

Validation of protocol

This protocol or parts of it has been used and validated in the following research article(s):

• Shi et al. [11] LFQRatio: A normalization method to decipher quantitative proteome changes in microbial coculture systems. Journal of Proteome Research (Figure 9).

• Kratzl et al. [21] Pseudomonas putida as saviour for troubled Synechococcus elongatus in a synthetic co-culture–interaction studies based on a multi-OMICs approach. Communications Biology (Figure 4).

General notes and troubleshooting

1. Growth details for S. elongatus and A. vinelandii: In our model system, we performed a time-course experiment to examine the growth and physiology of our S. elongatus and A. vinelandii co-culture over 8 days, harvesting samples for proteomics analysis on Days 0, 4, and 8. We monitored the growth both by optical density (OD) at 750 nm, which did not distinguish between strains but showed overall growth of the system, and by i) flow cytometry for S. elongatus and ii) calculating colony forming units for A. vinelandii. For more information on media, starting cell ratios, growth monitoring, and maintenance of these strains, we direct the reader to our paper [11].

2. Alternative in-browser programs for analyzing differentially expressed proteins: LFQ-Analyst was suggested here as an accessible, in-browser option for performing differential expression analysis of LFQ proteomics data [20]. Since performing our initial analysis [11], a developer version of LFQ-Analyst has been released (https://analyst-suites.org/apps/lfq-analyst-dev/), which allows for more control over parameter settings for differential expression analysis, as well as presence/absence analysis. Another powerful in-browser tool that we recommend is amica, which provides the user with an in-depth quality control analysis and parameter selection [22].

3. Dealing with missing values: Missing values in LFQ proteomics datasets are to be expected [23]. They can be broadly categorized as either “missing not at random” (MNAR), where the protein is either not expressed or has an abundance lower than the mass spectrometer’s limit of detection (i.e., biological reasons), or “missing at random” (MAR), where the protein is not identified for reasons related to the protocol and instrumentation (for more detail, see Jin et al. [24]). So as not to lose valuable data, we can apply various imputation algorithms to fill in these missing values.

We observed that missing values were particularly prevalent in mixed co-culture samples compared to their mono-culture counterparts, and the extent of this will likely differ between co-culture types. In our example application of LFQRatio, we applied the default imputation settings (Perseus type) for our DEP analysis [11]; we also applied this same imputation method in another study [21] in which we achieved a good correlation of the proteomics data with transcriptomics data, suggesting that this imputation method was acceptable. However, we advise researchers to consider which imputation method is most appropriate for their own dataset; for more information and guidance, we refer the reader to this paper: https://www.nature.com/articles/s41598-021-81279-4.

4. LFQ-Analyst details: LFQ-Analyst analyzes differentially expressed proteins according to the following steps: Filter out contaminant proteins, reverse sequences, and proteins identified “only by site;” remove proteins that have been only identified by a single peptide and proteins not identified/quantified consistently in the same condition; transform the LFQ data to log2 scale; group samples by conditions and missing values imputed using the selected imputation type [here, we used Perseus-type imputation, a MNAR method that takes random draws from a left-shifted Gaussian distribution of 1.8 StDev (standard deviation) apart with a width of 0.3]; use protein-wise linear models combined with empirical Bayes statistics for the differential expression analyses; and use the limma package from R Bioconductor to generate a list of differentially expressed proteins for each pair-wise comparison [20].

Supplementary information

The following supporting information can be downloaded here:

1. Table S1. Parameter settings of MaxQuant software used for protein identification and quantification.

2. File S1. R script for analyzing the pI of S. elongatus or A. vinelandii proteome.

3. File S2. R script for analyzing the MW of S. elongatus or A. vinelandii proteome.

4. File S3. R script for analyzing the hydrophobicity of S. elongatus or A. vinelandii proteome.

5. File S4. R script for analyzing shared peptides between S. elongatus and A. vinelandii.

Note: These scripts can be used for other organisms by changing the “sequence.fasta” file to the corresponding FASTA files.

6. Dataset S1. Protein sequences of S. elongatus cscB/SPS.

7. Dataset S2. Protein sequences of A. vinelandii ΔnifL.

8. Dataset S3. Protein sequences of S. elongatus cscB/SPS and A. vinelandii ΔnifL co-culture (merged database).

Acknowledgments

The authors acknowledge the Chinese Scholarship Council for funding a PhD studentship (201808370169) and EPSRC funding Engineering Microbial Consortia for Industry (EP/S020705/1). The QExactive HF orbitrap mass spectrometer was funded by BBSRC UK (award no. BB/M012166/1). This protocol was used in [11,21].

Competing interests

The authors declare no conflicts of interest.

References

1. Minty, J. J., Singer, M. E., Scholz, S. A., Bae, C. H., Ahn, J. H., Foster, C. E., Liao, J. C. and Lin, X. N. (2013). Design and characterization of synthetic fungal-bacterial consortia for direct production of isobutanol from cellulosic biomass. Proc Natl Acad Sci USA. 110(36): 14592–14597. https://doi.org/10.1073/pnas.1218447110
2. Pandhal, J. and Noirel, J. (2014). Synthetic microbial ecosystems for biotechnology. Biotechnol Lett. 36(6): 1141–1151. https://doi.org/10.1007/s10529-014-1480-y
3. Christie-Oleza, J. A., Sousoni, D., Lloyd, M., Armengaud, J. and Scanlan, D. J. (2017). Nutrient recycling facilitates long-term stability of marine microbial phototroph–heterotroph interactions. Nat Microbiol. 2(9): e100. https://doi.org/10.1038/nmicrobiol.2017.100
4. Ma, Q., Zhou, J., Zhang, W., Meng, X., Sun, J. and Yuan, Y. J. (2011). Integrated Proteomic and Metabolomic Analysis of an Artificial Microbial Community for Two-Step Production of Vitamin C. PLoS One. 6(10): e26108. https://doi.org/10.1371/journal.pone.0026108
5. Chignell, J. F., Park, S., Lacerda, C. M. R., De Long, S. K. and Reardon, K. F. (2018). Label-Free Proteomics of a Defined, Binary Co-culture Reveals Diversity of Competitive Responses Between Members of a Model Soil Microbial System. Microb Ecol. 75(3): 701–719. https://doi.org/10.1007/s00248-017-1072-1
6. Cappadona, S., Baker, P. R., Cutillas, P. R., Heck, A. J. R. and van Breukelen, B. (2012). Current challenges in software solutions for mass spectrometry-based quantitative proteomics. Amino Acids. 43(3): 1087–1108. https://doi.org/10.1007/s00726-012-1289-8
7. Cox, J., Hein, M. Y., Luber, C. A., Paron, I., Nagaraj, N. and Mann, M. (2014). Accurate Proteome-wide Label-free Quantification by Delayed Normalization and Maximal Peptide Ratio Extraction, Termed MaxLFQ. Mol Cell Proteomics. 13(9): 2513–2526. https://doi.org/10.1074/mcp.m113.031591
8. Ammar, C., Schessner, J. P., Willems, S., Michaelis, A. C. and Mann, M. (2023). Accurate Label-Free Quantification by directLFQ to Compare Unlimited Numbers of Proteomes. Mol Cell Proteomics. 22(7): 100581. https://doi.org/10.1016/j.mcpro.2023.100581
9. Kleiner, M. (2019). Metaproteomics: Much More than Measuring Gene Expression in Microbial Communities. mSystems. 4(3): e00115–19. https://doi.org/10.1128/msystems.00115-19
10. Goeminne, L. J., Gevaert, K. and Clement, L. (2018). Experimental design and data-analysis in label-free quantitative LC/MS proteomics: A tutorial with MSqRob. J Proteomics. 171: 23–36. https://doi.org/10.1016/j.jprot.2017.04.004
11. Shi, M., Evans, C. A., McQuillan, J. L., Noirel, J. and Pandhal, J. (2024). LFQRatio: A Normalization Method to Decipher Quantitative Proteome Changes in Microbial Coculture Systems. J Proteome Res. 23(3): 999–1013. https://doi.org/10.1021/acs.jproteome.3c00714
12. Abramson, B. W., Kachel, B., Kramer, D. M. and Ducat, D. C. (2016). Increased Photochemical Efficiency in Cyanobacteria via an Engineered Sucrose Sink. Plant Cell Physiol. 57(12): 2451–2460. https://doi.org/10.1093/pcp/pcw169
13. Ortiz-Marquez, J. C. F., Do Nascimento, M., Dublan, M. d. l. A. and Curatti, L. (2012). Association with an Ammonium-Excreting Bacterium Allows Diazotrophic Culture of Oil-Rich Eukaryotic Microalgae. Appl Environ Microbiol. 78(7): 2345–2352. https://doi.org/10.1128/aem.06260-11
14. Kozlowski, L. P. (2016). Proteome-pI: proteome isoelectric point database. Nucleic Acids Res. 45: D1112–D1116. https://doi.org/10.1093/nar/gkw978
15. Angel, T. E., Aryal, U. K., Hengel, S. M., Baker, E. S., Kelly, R. T., Robinson, E. W. and Smith, R. D. (2012). Mass spectrometry-based proteomics: existing capabilities and future directions. Chem Soc Rev. 41(10): 3912–3928. https://doi.org/10.1039/c2cs15331a
16. Warwood, S., Byron, A., Humphries, M. J. and Knight, D. (2013). The effect of peptide adsorption on signal linearity and a simple approach to improve reliability of quantification. J Proteomics. 85: 160–164. https://doi.org/10.1016/j.jprot.2013.04.034
17. Zubarev, R. A. (2013). The challenge of the proteome dynamic range and its implications for in-depth proteomics. Proteomics. 13(5): 723–726. https://doi.org/10.1002/pmic.201200451
18. Kumar, D., Yadav, A. K. and Dash, D. (2017). Choosing an Optimal Database for Protein Identification from Tandem Mass Spectrometry Data. In: Keerthikumar, S. and Mathivanan, S. (eds). Proteome Bioinformatics. 17–29. https://doi.org/10.1007/978-1-4939-6740-7_3
19. Helliwell, K. E., Pandhal, J., Cooper, M. B., Longworth, J., Kudahl, U. J., Russo, D. A., Tomsett, E. V., Bunbury, F., Salmon, D. L., Smirnoff, N., et al. (2017). Quantitative proteomics of a B₁₂‐dependent alga grown in coculture with bacteria reveals metabolic tradeoffs required for mutualism. New Phytol. 217(2): 599–612. https://doi.org/10.1111/nph.14832
20. Shah, A. D., Goode, R. J. A., Huang, C., Powell, D. R. and Schittenhelm, R. B. (2019). LFQ-Analyst: An Easy-To-Use Interactive Web Platform To Analyze and Visualize Label-Free Proteomics Data Preprocessed with MaxQuant. J Proteome Res. 19(1): 204–211. https://doi.org/10.1021/acs.jproteome.9b00496
21. Kratzl, F., Urban, M., Pandhal, J., Shi, M., Meng, C., Kleigrewe, K., Kremling, A. and Pflüger-Grau, K. (2024). Pseudomonas putida as saviour for troubled Synechococcus elongatus in a synthetic co-culture – interaction studies based on a multi-OMICs approach. Commun Biol. 7(1): 452. https://doi.org/10.1038/s42003-024-06098-5
22. Didusch, S., Madern, M., Hartl, M. and Baccarini, M. (2022). amica: an interactive and user-friendly web-platform for the analysis of proteomics data. BMC Genomics. 23: 817. https://doi.org/10.1101/2021.11.23.466958
23. Liu, M. and Dongre, A. (2020). Proper imputation of missing values in proteomics datasets for differential expression analysis. Briefings Bioinf. 22(3): 1–17. https://doi.org/10.1093/bib/bbaa112
24. Jin, L., Bi, Y., Hu, C., Qu, J., Shen, S., Wang, X. and Tian, Y. (2021). A comparative study of evaluating missing value imputation methods in label-free proteomics. Sci Rep. 11(1): 1760. https://doi.org/10.1038/s41598-021-81279-4