Update double_similar/README.md

double_similar/README.md

@@ -35,7 +35,7 @@ This pipeline identifies high-confidence anti-CRISPR (ACR) candidates from Folds
 │  Step 2: Structural Similarity Screening (Foldseek)             │
 │  ├─ 2a: Inhibition Type Probability Screening                   │
 │  │   - inhibition_type_probability = 1.0                        │
-│  │                                                             │
+│  │                                                              │
 │  └─ 2b: TM-score Thresholding                                   │
 │      - alntmscore > 0.7                                         │
 │      - qtmscore > 0.7                                           │
@@ -61,18 +61,14 @@ This pipeline identifies high-confidence anti-CRISPR (ACR) candidates from Folds
               ┌────────────┴────────────┐
               ▼                         ▼
        ┌──────────────┐          ┌──────────────┐
-       │  Pass: 12    │          │  Reject: 14  │
+       │  Pass: 10    │          │  Reject: 16  │
        │  records     │          │  (In DB)     │
        └──────┬───────┘          └──────────────┘
               │
               ▼
 ┌─────────────────────────────────────────────────────────────────┐
 │  Final Candidates                                               │
-│  - 12 novel high-confidence ACRs                                │
-│  - Distribution: AcrIF3(2), AcrIF11(2), AcrIE8.1(1),            │
-│                  AcrIB2(1), AcrIE2(1), AcrIIA12(1),             │
-│                  AcrIC9(1), AcrIIA14(1), AcrIE9(1),             │
-│                  AcrIC6(1)                                      │
+│  - 10 novel high-confidence ACRs                                │
 └─────────────────────────────────────────────────────────────────┘
 ```
 
@@ -90,14 +86,6 @@ This pipeline identifies high-confidence anti-CRISPR (ACR) candidates from Folds
 | `acr_verified_2025_pdb/*.pdb` | Reference structure database for Foldseek | anti-CRISPRdb v3.0 (verified) | - |
 | `anti-CRISPRdb3_for_psiblast.fasta` | Deduplication reference DB | anti-CRISPRdb v3.0 (non-predict) | 1,192 |
 
-### Output Files
-
-| File | Description | Records |
-|------|-------------|---------|
-| `seq_similar_foldseek_out_2025_acr_2026.csv` | Foldseek structural alignment results | 312 |
-| `filtered_high_confidence.csv` | High-confidence candidates (Steps 1-2) | 26 |
-| `filtered_not_in_db.csv` | Final novel candidates (All steps) | 12 |
-
 ---
 
 ## Software and Tools
@@ -125,7 +113,6 @@ Candidate prophage proteins are subjected to homology screening against the anti
 | Metric | Value |
 |--------|-------|
 | Input records | All prophage proteins |
-| Records passing | 312 |
 
 ### Step 2: Structural Similarity Screening (Foldseek-based)
 
@@ -137,11 +124,6 @@ Sequence-similar candidates from Step 1 undergo three-dimensional structural com
 
 This threshold ensures only candidates with the highest confidence for CRISPR-Cas inhibition mechanism are retained. The inhibition type probability represents the classifier's confidence that the query protein shares the same inhibition mechanism as the target ACR.
 
-| Metric | Value |
-|--------|-------|
-| Input records | 312 |
-| Records passing | 26 |
-| Retention rate | 8.3% |
 
 #### Step 2b: Structural Similarity Thresholding (TM-score)
 
@@ -155,81 +137,10 @@ This threshold ensures only candidates with the highest confidence for CRISPR-Ca
 
 **Rationale:** TM-score thresholds of >0.7 indicate proteins with highly similar 3D structures, suggesting conserved functional mechanisms. All three metrics are required to ensure bidirectional structural conservation between query and target.
 
-| Metric | Value |
-|--------|-------|
-| Input records | 26 |
-| Records passing | 26 |
-| Retention rate | 100% |
-
 ### Step 3: Reference Database Filtering
 
-**Database:** `anti-CRISPRdb3_for_psiblast.fasta`
+**Database:** `anti-CRISPRdb3（verified）`
 - Total sequences: 1,192
 - Source: anti-CRISPRdb v3.0 (excluding predicted entries)
 
-**Criterion:** Candidate sequence must **NOT** be present in the reference database.
-
-| Metric | Value |
-|--------|-------|
-| Input records | 26 |
-| Records removed (in DB) | 14 |
-| Records retained | 12 |
-| Retention rate | 46.2% |
-
----
-
-## Statistics Summary
-
-| Filtering Step | Records In | Records Out | Retention Rate | Cumulative Rate |
-|----------------|------------|-------------|----------------|-----------------|
-| Step 1: BLASTP screening | All | 312 | - | - |
-| Step 2a: Probability = 1.0 | 312 | 26 | 8.3% | 8.3% |
-| Step 2b: TM-scores > 0.7 | 26 | 26 | 100% | 8.3% |
-| Step 3: Remove DB matches | 26 | 12 | 46.2% | 3.8% |
-| **Final candidates** | **312** | **12** | **-** | **3.8%** |
-
----
-
-## Final ACR Candidate Distribution
-
-| ACR Type | Count | CRISPR-Cas System | Query IDs |
-|----------|-------|-------------------|-----------|
-| AcrIF3 | 2 | Type I-F | KTK51907.1, ESR95706.1 |
-| AcrIF11 | 2 | Type I-F | WP_057390551.1, WP_103182539.1 |
-| AcrIE8.1 | 1 | Type I-E | ATM37883.1 |
-| AcrIB2 | 1 | Type I-B | PBF40835.1 |
-| AcrIE2 | 1 | Type I-E | PBX41287.1 |
-| AcrIIA12 | 1 | Type II-A | PCZ07134.1 |
-| AcrIC9 | 1 | Type I-C | WP_016071148.1 |
-| AcrIIA14 | 1 | Type II-A | WP_411756513.1 |
-| AcrIE9 | 1 | Type I-E | WP_108234913.1 |
-| AcrIC6 | 1 | Type I-C | WP_108234914.1 |
-| **Total** | **12** | - | - |
-
----
-
-## Data Availability
-
-All input and output files are available in the `double_similar/` directory:
-
-- **Raw data:** `seq_similar_foldseek_out_2025_acr_2026.csv`
-- **Intermediate results:** `filtered_high_confidence.csv`
-- **Final candidates:** `filtered_not_in_db.csv`
-- **Reference database:** `anti-CRISPRdb3_for_psiblast.fasta`
-
----
-
-## References
-
-1. van Kempen, M. et al. Fast and accurate protein structure search with Foldseek. *Nat. Biotechnol.* **42**, 243-246 (2024).
-2. [anti-CRISPRdb reference to be added]
-
----
-
-## Contact
-
-For questions regarding this pipeline, please contact [email to be added].
-
----
-
-*Last updated: 2026-04-04*
+**Criterion:** Candidate sequence must **NOT** be present in the reference database.