Thermo class using primer3-py for thermodynamic calcs to replace NtThal and melting.py #110

tfenne · 2025-01-15T00:00:45Z

Creates a this wrapper around primer3-py's _ThermoAnalysis allowing us to add typing, docs, etc. Then replaces NtThal and melting.py with the new implementation.

codecov · 2025-01-15T00:02:06Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.37%. Comparing base (cba8c74) to head (4b06255).
Report is 1 commits behind head on dev.

Additional details and impacted files

@@            Coverage Diff             @@
##              dev     #110      +/-   ##
==========================================
+ Coverage   97.32%   97.37%   +0.04%     
==========================================
  Files          21       20       -1     
  Lines        1642     1635       -7     
  Branches      195      188       -7     
==========================================
- Hits         1598     1592       -6     
+ Misses         26       24       -2     
- Partials       18       19       +1

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

nh13

looking good to me so far

nh13 · 2025-01-15T03:57:25Z

prymer/thermo.py

+    annealing_temp_c: Actual annealing temperature of the PCR reaction in Celsius
+    temp_c: Simulation temperature for structure prediction in Celsius
+    max_nn_length: Maximum length in bases for nearest-neighbor calcs
+    max_loop: maximum size of loops in bases when predicting structures


Suggested change

max_loop: maximum size of loops in bases when predicting structures

max_loop: Maximum size of loops in bases when predicting structures

nh13 · 2025-01-15T03:57:46Z

prymer/thermo.py

+    TM_METHOD_BRESLAUER: ClassVar[str] = "breslauer"
+    TM_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"
+
+    SALT_CORRECTION_METHOD_SCHILDKRAUT: ClassVar[str] = "schildkraut"
+    SALT_CORRECTION_METHOD_OWCZARZY: ClassVar[str] = "owczarzy"
+    SALT_CORRECTION_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"


Why not enums?

I thought about it ... mostly wanted to keep the number of classes/symbols down, and felt like this was an ok compromise. That said, I'd like to expose these in PrimerAndAmpliconParameters and ProbeParameters and at that point enums may make better sense.

nh13

I think I'd prefer enums in the Thermo class' namespace.

nh13 · 2025-01-15T15:58:50Z

prymer/api/picking.py

@@ -119,6 +119,8 @@ def build_primer_pairs(  # noqa: C901
            and those exceeding the maximum Tm will be discarded
        weights: the set of penalty weights
        fasta_path: the path to the FASTA file from which the amplicon sequence will be retrieved.
+        thermo: a Thermo instance for performing thermodynamic calculations including amplicon tm;


Does this work? It'd be nice to link to the class

Suggested change

thermo: a Thermo instance for performing thermodynamic calculations including amplicon tm;

thermo: a [`Thermo`][prymer.thermo] instance for performing thermodynamic calculations including amplicon tm;

nh13 · 2025-01-15T15:59:34Z

prymer/primer3/primer3_parameters.py

+
+

Not sure how this happened!

Suggested change

nh13 · 2025-01-15T16:02:24Z

prymer/thermo.py

+        result: ThermoResult = self._thermo.calc_heterodimer(bases1, bases2)
+        return float(result.tm)
+
+    def heterodimer_3p_anchored_tm(self, bases1: str, bases2: str) -> float:


why not a boolean called 3p_anchored in heterodimer_tm?

I think it's a coin flip. You could also ask "why not a single dimer_tm() that takes an optional second sequence. At the end of the day I think it's clearer to have separate methods with separate documentation that we can elaborate on.

prymer/thermo.py

nh13 · 2025-01-15T16:11:51Z

prymer/thermo.py

+    TM_METHOD_BRESLAUER: ClassVar[str] = "breslauer"
+    TM_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"
+
+    SALT_CORRECTION_METHOD_SCHILDKRAUT: ClassVar[str] = "schildkraut"
+    SALT_CORRECTION_METHOD_OWCZARZY: ClassVar[str] = "owczarzy"
+    SALT_CORRECTION_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"


I think you could define the enums at the class level, to keep the namespace from being polluted.

Also consider using StrEnum. We may want to use LowercaseStrEnum from strenum as StrEnum isn't in the standard library until 3.11.

Suggested change

TM_METHOD_BRESLAUER: ClassVar[str] = "breslauer"

TM_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"

SALT_CORRECTION_METHOD_SCHILDKRAUT: ClassVar[str] = "schildkraut"

SALT_CORRECTION_METHOD_OWCZARZY: ClassVar[str] = "owczarzy"

SALT_CORRECTION_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"

@enum.unique

class TmMethod(Enum):

BRESLAUER = "breslauer"

SANTALUCIA = "santalucia"

@enum.unique

class SaltCorrectionMethod(Enum):

SCHILDKRAUT = "schildkraut"

OWCZARZY = "owczarzy"

SANTALUCIA = "santalucia"

I'm going to defer this for now. I hear you .. but actually the way these same things get provided to primer3 during design is as integers (i.e. 0, 1, or 2). So I want to think about where these belong and how best to manage them when I circle back and add them there.

nh13 · 2025-01-15T16:12:19Z

prymer/thermo.py

+            the melting temperature of the most likely (lowest energy) 3' anchored heterodimer
+        """
+        result: ThermoResult = self._thermo.calc_end_stability(bases1, bases2)
+        return float(result.tm)


Not covered due to copy-paste typo below

tests/test_thermo.py

nh13 · 2025-01-15T16:13:16Z

tests/test_thermo.py

+    salt_methods = [
+        Thermo.SALT_CORRECTION_METHOD_SANTALUCIA,
+        Thermo.SALT_CORRECTION_METHOD_SCHILDKRAUT,
+        Thermo.SALT_CORRECTION_METHOD_OWCZARZY
+    ]
+
+    tm_methods = [
+        Thermo.TM_METHOD_SANTALUCIA,
+        Thermo.TM_METHOD_BRESLAUER
+    ]


enums would avoid this, just saying.

Co-authored-by: Nils Homer <[email protected]>

nh13

Added the enum work into #108, as I moderately want it

tfenne requested a review from nh13 as a code owner January 15, 2025 00:00

nh13 reviewed Jan 15, 2025

View reviewed changes

Base automatically changed from tf_reorg_phase1 to dev January 15, 2025 12:38

tfenne force-pushed the tf_ntthal2 branch from 0329dff to 712af77 Compare January 15, 2025 13:46

tfenne changed the title ~~Sketch of using primer3-py for thermodynamic calcs.~~ Thermo class using primer3-py for thermodynamic calcs to replace NtThal and melting.py Jan 15, 2025

tfenne requested a review from nh13 January 15, 2025 14:06

nh13 requested changes Jan 15, 2025

View reviewed changes

tfenne added 6 commits January 15, 2025 09:16

Sketch of using primer3-py for thermodynamic calcs.

ae5b91b

Added docs and basic tests.

46a10db

Updated a couple of values to match primer3 defaults.

f701af5

Remove old long sequence Tm code.

3998295

Remove some commented out code.

4795d53

Removed old NtThal and replaced usage with new Thermo class.

0955ba8

coderabbitai bot approved these changes Jan 15, 2025

View reviewed changes

tfenne and others added 2 commits January 15, 2025 09:28

Fixups from code review.

4fb4c25

Apply suggestions from code review

4abd9e5

Co-authored-by: Nils Homer <[email protected]>

nh13 mentioned this pull request Jan 15, 2025

Primer 4.0 TODO List #108

Open

9 tasks

post review fixup

4b06255

tfenne force-pushed the tf_ntthal2 branch from 7f5b98b to 4b06255 Compare January 15, 2025 16:32

nh13 approved these changes Jan 15, 2025

View reviewed changes

tfenne merged commit 4348446 into dev Jan 15, 2025
5 checks passed

tfenne deleted the tf_ntthal2 branch January 15, 2025 16:39

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Thermo class using primer3-py for thermodynamic calcs to replace NtThal and melting.py #110

Thermo class using primer3-py for thermodynamic calcs to replace NtThal and melting.py #110

tfenne commented Jan 15, 2025 •

edited

Loading

codecov bot commented Jan 15, 2025 •

edited

Loading

nh13 left a comment

nh13 Jan 15, 2025

nh13 Jan 15, 2025

tfenne Jan 15, 2025

nh13 left a comment

nh13 Jan 15, 2025

nh13 Jan 15, 2025

nh13 Jan 15, 2025

tfenne Jan 15, 2025

nh13 Jan 15, 2025

tfenne Jan 15, 2025

nh13 Jan 15, 2025

nh13 Jan 15, 2025

nh13 left a comment

	max_loop: maximum size of loops in bases when predicting structures
	max_loop: Maximum size of loops in bases when predicting structures

	thermo: a Thermo instance for performing thermodynamic calculations including amplicon tm;
	thermo: a [`Thermo`][prymer.thermo] instance for performing thermodynamic calculations including amplicon tm;

-    TM_METHOD_BRESLAUER: ClassVar[str] = "breslauer"
-    TM_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"
-    SALT_CORRECTION_METHOD_SCHILDKRAUT: ClassVar[str] = "schildkraut"
-    SALT_CORRECTION_METHOD_OWCZARZY: ClassVar[str] = "owczarzy"
-    SALT_CORRECTION_METHOD_SANTALUCIA: ClassVar[str] = "santalucia"
+    @enum.unique
+    class TmMethod(Enum):
+        BRESLAUER = "breslauer"
+        SANTALUCIA = "santalucia"
+    @enum.unique
+    class SaltCorrectionMethod(Enum):
+        SCHILDKRAUT = "schildkraut"
+        OWCZARZY = "owczarzy"
+        SANTALUCIA = "santalucia"

Thermo class using primer3-py for thermodynamic calcs to replace NtThal and melting.py #110

Thermo class using primer3-py for thermodynamic calcs to replace NtThal and melting.py #110

Conversation

tfenne commented Jan 15, 2025 • edited Loading

codecov bot commented Jan 15, 2025 • edited Loading

Codecov Report

nh13 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

nh13 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

nh13 left a comment

Choose a reason for hiding this comment

tfenne commented Jan 15, 2025 •

edited

Loading

codecov bot commented Jan 15, 2025 •

edited

Loading