This document explains the background and rationale of your RNA-seq dataset based on sample names, updates the genotype mapping (AB = ΔadeAB; IJ = ΔadeIJ; WT19606 = WT; W1/Y1 = clinical WT isolates), and provides ready-to-run analysis contrasts aligned with your research proposal.

---

1) Naming logic (decoded)

• Medium prefix

  • LB- → growth in LB (non-selective baseline; total transcriptome without membrane/efflux stress).
  • Mac- → growth in MacConkey (bile salts + crystal violet; membrane/outer-membrane/efflux stress condition).

• Strain/genotype tag

  • WT19606 → A. baumannii ATCC 19606, wild-type (reference WT).
  • IJ → ΔadeIJ (AdeIJK RND efflux knockout).
  • AB → ΔadeAB (AdeAB subfamily knockout).
  • W1, Y1 → two clinical wild-type isolates (distinct lineages).

• Replicates

  • r1 / r2 / r3 / r4 → biological replicates.

Note on proposal text: The provided proposal did not explicitly name W1/Y1; it referred to “clinical isolates” generally. If you analyze W1/Y1, add a Methods line defining them (e.g., source, genotype as WT).

---

2) Updated mapping table

Sample prefix	Medium	Strain	Genotype	Purpose / Rationale
**LB-AB-***	LB	AB	ΔadeAB	Baseline transcriptome of AdeAB knockout (no stress)
**Mac-AB-***	MacConkey	AB	ΔadeAB	Stress response without AdeAB (efflux-dependent programs)
**LB-IJ-***	LB	IJ	ΔadeIJ	Baseline of AdeIJK knockout
**Mac-IJ-***	MacConkey	IJ	ΔadeIJ	Stress response without AdeIJK
**LB-WT19606-***	LB	ATCC 19606	WT	Reference WT baseline (anchor strain)
**Mac-WT19606-***	MacConkey	ATCC 19606	WT	Reference WT under stress (gold-standard contrast vs ΔadeAB/ΔadeIJ)
**LB-W1-***	LB	W1	WT (clinical)	Baseline of clinical WT lineage W1
**Mac-W1-***	MacConkey	W1	WT (clinical)	W1 under stress
**LB-Y1-***	LB	Y1	WT (clinical)	Baseline of clinical WT lineage Y1
**Mac-Y1-***	MacConkey	Y1	WT (clinical)	Y1 under stress

Why these conditions matter (proposal-aligned):

• LB captures baseline networks.
• Mac induces the membrane/efflux stress program that revealed R vs S behavior in your proposal and is tightly linked to RND function.
• Contrasting WT vs efflux knockouts (ΔadeAB/ΔadeIJ) under LB/Mac tests both genotype main effects and genotype × stress interactions.
• Multiple WT lineages (WT19606/W1/Y1) allow testing conservation vs isolate-specificity of stress responses (avoid overfitting to one WT).

---

3) Suggested metadata file (samples.csv)

sample,fastq_1,fastq_2,medium,strain,genotype,replicate,notes,strandedness
LB-AB-r1,LB-AB-r1_R1.fq.gz,LB-AB-r1_R2.fq.gz,LB,AB,ΔadeAB,r1,AdeAB knockout baseline,auto
LB-AB-r2,LB-AB-r2_R1.fq.gz,LB-AB-r2_R2.fq.gz,LB,AB,ΔadeAB,r2,AdeAB knockout baseline,auto
LB-AB-r3,LB-AB-r3_R1.fq.gz,LB-AB-r3_R2.fq.gz,LB,AB,ΔadeAB,r3,AdeAB knockout baseline,auto
LB-IJ-r1,LB-IJ-r1_R1.fq.gz,LB-IJ-r1_R2.fq.gz,LB,IJ,ΔadeIJ,r1,AdeIJK knockout baseline,auto
LB-IJ-r2,LB-IJ-r2_R1.fq.gz,LB-IJ-r2_R2.fq.gz,LB,IJ,ΔadeIJ,r2,AdeIJK knockout baseline,auto
LB-IJ-r4,LB-IJ-r4_R1.fq.gz,LB-IJ-r4_R2.fq.gz,LB,IJ,ΔadeIJ,r4,AdeIJK knockout baseline,auto
LB-W1-r1,LB-W1-r1_R1.fq.gz,LB-W1-r1_R2.fq.gz,LB,W1,WT,r1,clinical WT W1 baseline,auto
LB-W1-r2,LB-W1-r2_R1.fq.gz,LB-W1-r2_R2.fq.gz,LB,W1,WT,r2,clinical WT W1 baseline,auto
LB-W1-r3,LB-W1-r3_R1.fq.gz,LB-W1-r3_R2.fq.gz,LB,W1,WT,r3,clinical WT W1 baseline,auto
LB-WT19606-r2,LB-WT19606-r2_R1.fq.gz,LB-WT19606-r2_R2.fq.gz,LB,WT19606,WT,r2,reference WT baseline,auto
LB-WT19606-r3,LB-WT19606-r3_R1.fq.gz,LB-WT19606-r3_R2.fq.gz,LB,WT19606,WT,r3,reference WT baseline,auto
LB-WT19606-r4,LB-WT19606-r4_R1.fq.gz,LB-WT19606-r4_R2.fq.gz,LB,WT19606,WT,r4,reference WT baseline,auto
LB-Y1-r2,LB-Y1-r2_R1.fq.gz,LB-Y1-r2_R2.fq.gz,LB,Y1,WT,r2,clinical WT Y1 baseline,auto
LB-Y1-r3,LB-Y1-r3_R1.fq.gz,LB-Y1-r3_R2.fq.gz,LB,Y1,WT,r3,clinical WT Y1 baseline,auto
LB-Y1-r4,LB-Y1-r4_R1.fq.gz,LB-Y1-r4_R2.fq.gz,LB,Y1,WT,r4,clinical WT Y1 baseline,auto
Mac-AB-r1,Mac-AB-r1_R1.fq.gz,Mac-AB-r1_R2.fq.gz,MacConkey,AB,ΔadeAB,r1,AdeAB knockout under stress,auto
Mac-AB-r2,Mac-AB-r2_R1.fq.gz,Mac-AB-r2_R2.fq.gz,MacConkey,AB,ΔadeAB,r2,AdeAB knockout under stress,auto
Mac-AB-r3,Mac-AB-r3_R1.fq.gz,Mac-AB-r3_R2.fq.gz,MacConkey,AB,ΔadeAB,r3,AdeAB knockout under stress,auto
Mac-IJ-r1,Mac-IJ-r1_R1.fq.gz,Mac-IJ-r1_R2.fq.gz,MacConkey,IJ,ΔadeIJ,r1,AdeIJK knockout under stress,auto
Mac-IJ-r2,Mac-IJ-r2_R1.fq.gz,Mac-IJ-r2_R2.fq.gz,MacConkey,IJ,ΔadeIJ,r2,AdeIJK knockout under stress,auto
Mac-IJ-r4,Mac-IJ-r4_R1.fq.gz,Mac-IJ-r4_R2.fq.gz,MacConkey,IJ,ΔadeIJ,r4,AdeIJK knockout under stress,auto
Mac-W1-r1,Mac-W1-r1_R1.fq.gz,Mac-W1-r1_R2.fq.gz,MacConkey,W1,WT,r1,clinical WT W1 under stress,auto
Mac-W1-r2,Mac-W1-r2_R1.fq.gz,Mac-W1-r2_R2.fq.gz,MacConkey,W1,WT,r2,clinical WT W1 under stress,auto
Mac-W1-r3,Mac-W1-r3_R1.fq.gz,Mac-W1-r3_R2.fq.gz,MacConkey,W1,WT,r3,clinical WT W1 under stress,auto
Mac-WT19606-r2,Mac-WT19606-r2_R1.fq.gz,Mac-WT19606-r2_R2.fq.gz,MacConkey,WT19606,WT,r2,reference WT under stress,auto
Mac-WT19606-r3,Mac-WT19606-r3_R1.fq.gz,Mac-WT19606-r3_R2.fq.gz,MacConkey,WT19606,WT,r3,reference WT under stress,auto
Mac-WT19606-r4,Mac-WT19606-r4_R1.fq.gz,Mac-WT19606-r4_R2.fq.gz,MacConkey,WT19606,WT,r4,reference WT under stress,auto
Mac-Y1-r2,Mac-Y1-r2_R1.fq.gz,Mac-Y1-r2_R2.fq.gz,MacConkey,Y1,WT,r2,clinical WT Y1 under stress,auto
Mac-Y1-r3,Mac-Y1-r3_R1.fq.gz,Mac-Y1-r3_R2.fq.gz,MacConkey,Y1,WT,r3,clinical WT Y1 under stress,auto
Mac-Y1-r4,Mac-Y1-r4_R1.fq.gz,Mac-Y1-r4_R2.fq.gz,MacConkey,Y1,WT,r4,clinical WT Y1 under stress,auto

---

4) DE design & contrasts (DESeq2/edgeR)

Model

• Reference-genotype focus (WT19606 vs knockouts):
  ~ medium * genotype
  where medium ∈ {LB, MacConkey}, genotype ∈ {WT, ΔadeAB, ΔadeIJ}.
  (Set WT19606_LB as reference level.)

• All strains (WT lineages):
  ~ medium * strain
  where strain ∈ {WT19606, W1, Y1, ΔadeAB, ΔadeIJ}.

Hypothesis-driven contrasts

1. Stress response within each background:
   Mac vs LB for WT19606, ΔadeAB, ΔadeIJ, W1, Y1.
   → Membrane/efflux-stress regulons; check adeA/B/C, adeI/J/K, envelope stress, OM biogenesis, osmotic/ions, ribosome PQC.

2. Genotype main effect at baseline (LB):
   ΔadeAB vs WT19606 (LB), ΔadeIJ vs WT19606 (LB).
   → Efflux-independent differences; compensatory pathways.

3. Genotype effect under stress (Mac):
   ΔadeAB vs WT19606 (Mac), ΔadeIJ vs WT19606 (Mac).
   → How loss of AdeAB/AdeIJK alters the stress transcriptome.

4. Interaction (genotype × medium):
   (ΔadeAB_Mac − ΔadeAB_LB) − (WT19606_Mac − WT19606_LB); same for ΔadeIJ.
   → Core test: does RND loss change the stress response itself?

5. Conservation across WT lineages:
   Intersect Mac vs LB DEGs across WT19606/W1/Y1 to define a conserved “Mac stress signature”; then identify isolate-specific modules.

QC

• Verify strandedness with RSeQC (you set auto), mapping rate, rRNA %, insert size.
• PCA: expect Mac vs LB separation; ΔadeIJ should diverge strongly under Mac.
• Validate a small panel via RT-qPCR (ade genes + OM stress markers).

---

5) Where this plugs into the proposal

• Mac vs LB embodies the proposal’s “LB = who is alive” vs “Mac = who can cope” paradigm (membrane/efflux stress).
• ΔadeAB / ΔadeIJ model the role of RND efflux in stress adaptation and heterogeneity.
• Multiple WT lineages prevent overfitting to ATCC 19606 and test generalizability.
• Downstream integration with PAP/MDK/R% metrics links transcriptome to phenotype (R vs S).

---

If you want, I can also generate a starter DESeq2 R script that reads this samples.csv, sets factors/contrasts, and outputs PCA, volcano plots, and KEGG/GO enrichment stubs.